Correlation between aggregated molecular cancer subtypes and selected clinical features

Overview

Introduction

This pipeline computes the correlation between cancer subtypes identified by different molecular patterns and selected clinical features.

Summary

Testing the association between subtypes identified by 10 different clustering approaches and 39 clinical features across 307 patients, 76 significant findings detected with P value < 0.05 and Q value < 0.25.

5 subtypes identified in current cancer cohort by 'Copy Number Ratio CNMF subtypes'. These subtypes correlate to 'YEARS_TO_BIRTH', 'PATHOLOGY_T_STAGE', 'RADIATION_THERAPY', 'HISTOLOGICAL_TYPE', 'NUMBER_OF_LYMPH_NODES', 'NEOPLASM_HISTOLOGIC_GRADE', 'LYMPH_NODES_EXAMINED_HE_COUNT', 'AGE_AT_DIAGNOSIS', and 'CLINICAL_STAGE'.
3 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. These subtypes correlate to 'HISTOLOGICAL_TYPE', 'NUMBER_OF_LYMPH_NODES', 'PREGNANCIES_COUNT_TOTAL', 'PREGNANCIES_COUNT_LIVE_BIRTH', 'LYMPH_NODES_EXAMINED_HE_COUNT', and 'KERATINIZATION_SQUAMOUS_CELL'.
CNMF clustering analysis on RPPA data identified 3 subtypes that correlate to 'LYMPHOVASCULAR_INVOLVEMENT' and 'LYMPH_NODES_EXAMINED'.
Consensus hierarchical clustering analysis on RPPA data identified 5 subtypes that correlate to 'HISTOLOGICAL_TYPE', 'KERATINIZATION_SQUAMOUS_CELL', and 'CHEMO_CONCURRENT_TYPE'.
CNMF clustering analysis on sequencing-based mRNA expression data identified 4 subtypes that correlate to 'YEARS_TO_BIRTH', 'HISTOLOGICAL_TYPE', 'NUMBER_OF_LYMPH_NODES', 'RACE', 'NEOPLASM_HISTOLOGIC_GRADE', 'PREGNANCIES_COUNT_TOTAL', 'PREGNANCIES_COUNT_LIVE_BIRTH', 'MENOPAUSE_STATUS', 'LYMPH_NODES_EXAMINED_HE_COUNT', 'KERATINIZATION_SQUAMOUS_CELL', 'AGE_AT_DIAGNOSIS', and 'CLINICAL_STAGE'.
Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'HISTOLOGICAL_TYPE', 'NUMBER_OF_LYMPH_NODES', 'PREGNANCIES_COUNT_TOTAL', 'PREGNANCIES_COUNT_LIVE_BIRTH', 'MENOPAUSE_STATUS', 'LYMPH_NODES_EXAMINED_HE_COUNT', 'KERATINIZATION_SQUAMOUS_CELL', 'INITIAL_PATHOLOGIC_DX_YEAR', 'HISTORY_HORMONAL_CONTRACEPTIVES_USE', and 'AGE_AT_DIAGNOSIS'.
3 subtypes identified in current cancer cohort by 'MIRSEQ CNMF'. These subtypes correlate to 'HISTOLOGICAL_TYPE', 'NEOPLASM_HISTOLOGIC_GRADE', 'TOBACCO_SMOKING_HISTORY', 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT', 'KERATINIZATION_SQUAMOUS_CELL', 'INITIAL_PATHOLOGIC_DX_YEAR', and 'CLINICAL_STAGE'.
5 subtypes identified in current cancer cohort by 'MIRSEQ CHIERARCHICAL'. These subtypes correlate to 'YEARS_TO_BIRTH', 'PATHOLOGY_T_STAGE', 'HISTOLOGICAL_TYPE', 'NEOPLASM_HISTOLOGIC_GRADE', 'PREGNANCIES_COUNT_LIVE_BIRTH', 'MENOPAUSE_STATUS', 'INITIAL_PATHOLOGIC_DX_YEAR', 'AGE_AT_DIAGNOSIS', and 'CLINICAL_STAGE'.
3 subtypes identified in current cancer cohort by 'MIRseq Mature CNMF subtypes'. These subtypes correlate to 'YEARS_TO_BIRTH', 'HISTOLOGICAL_TYPE', 'NEOPLASM_HISTOLOGIC_GRADE', 'MENOPAUSE_STATUS', 'KERATINIZATION_SQUAMOUS_CELL', 'INITIAL_PATHOLOGIC_DX_YEAR', 'AGE_AT_DIAGNOSIS', and 'CLINICAL_STAGE'.
5 subtypes identified in current cancer cohort by 'MIRseq Mature cHierClus subtypes'. These subtypes correlate to 'YEARS_TO_BIRTH', 'HISTOLOGICAL_TYPE', 'NUMBER_OF_LYMPH_NODES', 'NEOPLASM_HISTOLOGIC_GRADE', 'RADIATION_THERAPY_STATUS', 'LYMPH_NODES_EXAMINED_HE_COUNT', 'INITIAL_PATHOLOGIC_DX_YEAR', and 'AGE_AT_DIAGNOSIS'.

Results

Overview of the results

Table 1. Get Full Table Overview of the association between subtypes identified by 10 different clustering approaches and 39 clinical features. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, 76 significant findings detected.


Clinical Features	Statistical Tests	Copy Number Ratio CNMF subtypes	METHLYATION CNMF	RPPA CNMF subtypes	RPPA cHierClus subtypes	RNAseq CNMF subtypes	RNAseq cHierClus subtypes	MIRSEQ CNMF	MIRSEQ CHIERARCHICAL	MIRseq Mature CNMF subtypes	MIRseq Mature cHierClus subtypes
Time to Death	logrank test	0.682 (0.852)	0.967 (0.992)	0.0599 (0.285)	0.183 (0.498)	0.44 (0.723)	0.0302 (0.199)	0.889 (0.957)	0.692 (0.856)	0.779 (0.893)	0.928 (0.98)
YEARS TO BIRTH	Kruskal-Wallis (anova)	0.00931 (0.114)	0.174 (0.493)	0.0786 (0.325)	0.0879 (0.346)	8.51e-06 (0.000433)	0.00802 (0.114)	0.619 (0.844)	0.000125 (0.00376)	0.0366 (0.207)	0.00654 (0.102)
PATHOLOGY T STAGE	Fisher's exact test	0.0336 (0.199)	0.422 (0.72)	0.891 (0.957)	0.645 (0.844)	0.661 (0.844)	0.704 (0.857)	0.172 (0.493)	0.00365 (0.0729)	0.211 (0.541)	0.262 (0.567)
PATHOLOGY N STAGE	Fisher's exact test	0.35 (0.657)	0.232 (0.549)	0.661 (0.844)	0.213 (0.543)	0.377 (0.675)	0.349 (0.657)	0.609 (0.844)	0.23 (0.549)	0.547 (0.794)	0.0778 (0.325)
PATHOLOGY M STAGE	Fisher's exact test	0.736 (0.872)	0.36 (0.661)	0.55 (0.794)	1 (1.00)	0.267 (0.576)	0.0524 (0.266)	0.0987 (0.37)	0.354 (0.66)	0.114 (0.398)	0.282 (0.601)
RADIATION THERAPY	Fisher's exact test	0.0136 (0.144)	0.773 (0.891)	0.611 (0.844)	0.303 (0.624)	0.888 (0.957)	1 (1.00)	0.955 (0.986)	0.445 (0.726)	0.976 (1)	0.927 (0.98)
HISTOLOGICAL TYPE	Fisher's exact test	1e-05 (0.000433)	1e-05 (0.000433)	0.311 (0.628)	0.00394 (0.0732)	1e-05 (0.000433)	1e-05 (0.000433)	1e-05 (0.000433)	1e-05 (0.000433)	1e-05 (0.000433)	1e-05 (0.000433)
NUMBER PACK YEARS SMOKED	Kruskal-Wallis (anova)	0.662 (0.844)	0.112 (0.398)	0.198 (0.515)	0.54 (0.794)	0.106 (0.385)	0.437 (0.723)	0.0659 (0.29)	0.157 (0.474)	0.302 (0.624)	0.0646 (0.29)
NUMBER OF LYMPH NODES	Kruskal-Wallis (anova)	0.048 (0.247)	0.0366 (0.207)	0.632 (0.844)	0.177 (0.493)	0.0146 (0.146)	0.0308 (0.199)	0.242 (0.549)	0.134 (0.421)	0.478 (0.742)	0.0318 (0.199)
RACE	Fisher's exact test	0.239 (0.549)	0.273 (0.586)	0.307 (0.628)	0.209 (0.539)	0.00901 (0.114)	0.307 (0.628)	0.217 (0.547)	0.33 (0.647)	0.0823 (0.327)	0.0739 (0.313)
ETHNICITY	Fisher's exact test	0.225 (0.549)	0.701 (0.857)	0.432 (0.723)	0.473 (0.742)	0.255 (0.566)	0.896 (0.957)	0.674 (0.848)	0.412 (0.715)	0.24 (0.549)	0.217 (0.547)
WEIGHT KG AT DIAGNOSIS	Kruskal-Wallis (anova)	0.324 (0.638)	0.244 (0.55)	0.0556 (0.278)	0.542 (0.794)	0.694 (0.856)	0.677 (0.849)	0.551 (0.794)	0.933 (0.984)	0.476 (0.742)	0.479 (0.742)
TUMOR STATUS	Fisher's exact test	0.949 (0.986)	0.561 (0.801)	0.674 (0.848)	0.53 (0.789)	0.38 (0.676)	0.425 (0.72)	0.58 (0.814)	0.713 (0.861)	0.556 (0.797)	0.642 (0.844)
NEOPLASM HISTOLOGIC GRADE	Fisher's exact test	0.0298 (0.199)	0.375 (0.675)	0.34 (0.65)	0.551 (0.794)	0.00049 (0.0136)	0.157 (0.474)	0.00417 (0.0739)	0.0228 (0.199)	0.00354 (0.0729)	6e-05 (0.00213)
TOBACCO SMOKING YEAR STOPPED	Kruskal-Wallis (anova)	0.401 (0.698)	0.497 (0.759)	0.838 (0.93)	0.706 (0.857)	0.836 (0.93)	0.569 (0.81)	0.125 (0.415)	0.147 (0.455)	0.422 (0.72)	0.479 (0.742)
TOBACCO SMOKING PACK YEARS SMOKED	Kruskal-Wallis (anova)	0.662 (0.844)	0.112 (0.398)	0.198 (0.515)	0.54 (0.794)	0.106 (0.385)	0.437 (0.723)	0.0659 (0.29)	0.157 (0.474)	0.302 (0.624)	0.0646 (0.29)
TOBACCO SMOKING HISTORY	Kruskal-Wallis (anova)	0.338 (0.65)	0.0617 (0.29)	0.802 (0.91)	0.471 (0.742)	0.255 (0.566)	0.0944 (0.364)	0.043 (0.233)	0.669 (0.848)	0.444 (0.726)	0.31 (0.628)
AGEBEGANSMOKINGINYEARS	Kruskal-Wallis (anova)	0.527 (0.787)	0.715 (0.861)	0.747 (0.88)	0.808 (0.91)	0.226 (0.549)	0.655 (0.844)	0.892 (0.957)	0.764 (0.89)	0.95 (0.986)	0.4 (0.698)
RADIATION THERAPY STATUS	Fisher's exact test	1 (1.00)	1 (1.00)	0.725 (0.87)	1 (1.00)	0.733 (0.871)	1 (1.00)	0.551 (0.794)	0.858 (0.935)	1 (1.00)	0.0332 (0.199)
PREGNANCIES COUNT TOTAL	Kruskal-Wallis (anova)	0.669 (0.848)	0.0168 (0.16)	0.472 (0.742)	0.622 (0.844)	0.0261 (0.199)	0.0475 (0.247)	0.506 (0.771)	0.0673 (0.292)	0.494 (0.759)	0.0979 (0.37)
PREGNANCIES COUNT STILLBIRTH	Kruskal-Wallis (anova)	0.0563 (0.278)	0.793 (0.904)	0.458 (0.737)	0.224 (0.549)	0.164 (0.487)	0.574 (0.81)	0.756 (0.886)	0.481 (0.742)	0.756 (0.886)	0.657 (0.844)
PREGNANCY SPONTANEOUS ABORTION COUNT	Kruskal-Wallis (anova)	0.951 (0.986)	0.629 (0.844)	0.806 (0.91)	0.809 (0.91)	0.772 (0.891)	0.13 (0.415)	0.0206 (0.187)	0.179 (0.493)	0.194 (0.512)	0.159 (0.477)
PREGNANCIES COUNT LIVE BIRTH	Kruskal-Wallis (anova)	0.26 (0.567)	0.0333 (0.199)	0.124 (0.415)	0.114 (0.398)	0.0381 (0.209)	0.0188 (0.174)	0.649 (0.844)	0.016 (0.156)	0.957 (0.986)	0.235 (0.549)
PREGNANCY THERAPEUTIC ABORTION COUNT	Kruskal-Wallis (anova)	0.854 (0.935)	0.76 (0.887)	0.786 (0.898)	0.655 (0.844)	0.815 (0.913)	0.616 (0.844)	0.462 (0.739)	0.936 (0.984)	0.188 (0.506)	0.323 (0.638)
PREGNANCIES COUNT ECTOPIC	Kruskal-Wallis (anova)	0.6 (0.839)	0.291 (0.61)	0.919 (0.976)	0.638 (0.844)	0.192 (0.51)	0.855 (0.935)	0.338 (0.65)	0.958 (0.986)	0.982 (1.00)	0.426 (0.72)
POS LYMPH NODE LOCATION	Fisher's exact test	0.177 (0.493)	0.366 (0.666)	0.366 (0.666)	0.251 (0.562)	0.359 (0.661)	0.426 (0.72)	0.0901 (0.352)	0.428 (0.72)	0.397 (0.697)	0.261 (0.567)
MENOPAUSE STATUS	Fisher's exact test	0.241 (0.549)	0.0814 (0.327)	0.128 (0.415)	0.873 (0.949)	0.0481 (0.247)	0.0234 (0.199)	0.727 (0.87)	0.0134 (0.144)	0.012 (0.137)	0.29 (0.61)
LYMPHOVASCULAR INVOLVEMENT	Fisher's exact test	0.172 (0.493)	0.224 (0.549)	0.0301 (0.199)	0.425 (0.72)	0.819 (0.915)	0.639 (0.844)	0.607 (0.844)	0.378 (0.675)	0.695 (0.856)	0.112 (0.398)
LYMPH NODES EXAMINED HE COUNT	Kruskal-Wallis (anova)	0.048 (0.247)	0.0366 (0.207)	0.632 (0.844)	0.177 (0.493)	0.0146 (0.146)	0.0308 (0.199)	0.242 (0.549)	0.134 (0.421)	0.478 (0.742)	0.0318 (0.199)
LYMPH NODES EXAMINED	Kruskal-Wallis (anova)	0.742 (0.877)	0.804 (0.91)	0.00935 (0.114)	0.236 (0.549)	0.174 (0.493)	0.342 (0.65)	0.847 (0.933)	0.319 (0.635)	0.642 (0.844)	0.316 (0.634)
KERATINIZATION SQUAMOUS CELL	Fisher's exact test	0.524 (0.786)	0.0271 (0.199)	0.0795 (0.325)	0.0376 (0.209)	0.0321 (0.199)	0.00374 (0.0729)	0.0335 (0.199)	0.123 (0.415)	0.00884 (0.114)	0.0641 (0.29)
INITIAL PATHOLOGIC DX YEAR	Kruskal-Wallis (anova)	0.63 (0.844)	0.947 (0.986)	0.524 (0.786)	0.841 (0.93)	0.347 (0.657)	0.0135 (0.144)	0.000899 (0.0219)	0.0235 (0.199)	0.00872 (0.114)	0.0245 (0.199)
HISTORY HORMONAL CONTRACEPTIVES USE	Fisher's exact test	0.776 (0.892)	0.235 (0.549)	0.697 (0.856)	0.552 (0.794)	0.0953 (0.364)	0.0336 (0.199)	0.731 (0.871)	0.317 (0.634)	0.625 (0.844)	0.662 (0.844)
HEIGHT CM AT DIAGNOSIS	Kruskal-Wallis (anova)	0.461 (0.739)	0.573 (0.81)	0.126 (0.415)	0.167 (0.493)	0.291 (0.61)	0.146 (0.455)	0.904 (0.963)	0.686 (0.854)	0.708 (0.857)	0.437 (0.723)
CORPUS INVOLVEMENT	Fisher's exact test	0.95 (0.986)	0.17 (0.493)	0.127 (0.415)	0.455 (0.737)	0.0691 (0.296)	0.342 (0.65)	0.577 (0.812)	0.339 (0.65)	0.857 (0.935)	0.359 (0.661)
CHEMO CONCURRENT TYPE	Fisher's exact test	0.259 (0.567)	0.628 (0.844)	0.509 (0.772)	0.00359 (0.0729)	0.396 (0.697)	0.117 (0.402)	0.841 (0.93)	0.458 (0.737)	0.895 (0.957)	0.08 (0.325)
CERVIX SUV RESULTS	Kruskal-Wallis (anova)	0.768 (0.891)	0.385 (0.683)	0.0662 (0.29)		0.231 (0.549)	0.226 (0.549)	0.367 (0.666)	0.242 (0.549)	0.18 (0.493)	0.19 (0.509)
AGE AT DIAGNOSIS	Kruskal-Wallis (anova)	0.0112 (0.132)	0.184 (0.498)	0.0597 (0.285)	0.117 (0.402)	1.34e-05 (0.000522)	0.00655 (0.102)	0.698 (0.856)	9.35e-05 (0.00304)	0.0326 (0.199)	0.00599 (0.102)
CLINICAL STAGE	Fisher's exact test	0.0246 (0.199)	0.0591 (0.285)	0.645 (0.844)	0.13 (0.415)	0.0304 (0.199)	0.106 (0.385)	0.00923 (0.114)	0.00075 (0.0195)	0.0295 (0.199)	0.514 (0.778)

Clustering Approach #1: 'Copy Number Ratio CNMF subtypes'

Table S1. Description of clustering approach #1: 'Copy Number Ratio CNMF subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	15	112	56	59	53

'Copy Number Ratio CNMF subtypes' versus 'Time to Death'

P value = 0.682 (logrank test), Q value = 0.85

Table S2. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	282	68	0.1 - 210.7 (22.6)
subtype1	14	1	1.4 - 101.8 (26.4)
subtype2	108	29	0.4 - 155.8 (20.7)
subtype3	53	11	0.2 - 210.7 (23.0)
subtype4	58	15	0.1 - 177.0 (22.8)
subtype5	49	12	0.1 - 173.3 (25.0)

Figure S1. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'Copy Number Ratio CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.00931 (Kruskal-Wallis (anova)), Q value = 0.11

Table S3. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	293	48.1 (13.8)
subtype1	15	45.8 (12.3)
subtype2	111	45.6 (14.0)
subtype3	56	53.4 (12.4)
subtype4	59	48.0 (12.8)
subtype5	52	48.5 (15.2)

Figure S2. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.0336 (Fisher's exact test), Q value = 0.2

Table S4. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	135	70	21	8
subtype1	3	8	2	0
subtype2	50	25	6	2
subtype3	21	12	8	4
subtype4	32	17	2	1
subtype5	29	8	3	1

Figure S3. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.35 (Fisher's exact test), Q value = 0.66

Table S5. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	129	56
subtype1	7	3
subtype2	42	22
subtype3	19	13
subtype4	34	9
subtype5	27	9

Figure S4. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.736 (Fisher's exact test), Q value = 0.87

Table S6. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	108	10
subtype1	5	0
subtype2	42	4
subtype3	22	1
subtype4	22	4
subtype5	17	1

Figure S5. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.0136 (Fisher's exact test), Q value = 0.14

Table S7. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	53	125
subtype1	1	8
subtype2	15	52
subtype3	8	31
subtype4	17	19
subtype5	12	15

Figure S6. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

'Copy Number Ratio CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 1e-05 (Fisher's exact test), Q value = 0.00043

Table S8. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	6	242	6	21	3	17
subtype1	1	14	0	0	0	0
subtype2	3	96	2	3	1	7
subtype3	0	56	0	0	0	0
subtype4	2	31	4	13	2	7
subtype5	0	45	0	5	0	3

Figure S7. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'Copy Number Ratio CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.662 (Kruskal-Wallis (anova)), Q value = 0.84

Table S9. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	90	17.5 (14.3)
subtype1	6	14.2 (10.2)
subtype2	32	16.7 (12.2)
subtype3	20	22.4 (17.9)
subtype4	14	16.0 (10.8)
subtype5	18	15.7 (16.6)

Figure S8. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'Copy Number Ratio CNMF subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.048 (Kruskal-Wallis (anova)), Q value = 0.25

Table S10. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	150	0.9 (2.1)
subtype1	6	0.3 (0.8)
subtype2	48	0.8 (1.3)
subtype3	23	1.7 (2.3)
subtype4	41	0.9 (2.8)
subtype5	32	0.8 (1.8)

Figure S9. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'Copy Number Ratio CNMF subtypes' versus 'RACE'

P value = 0.239 (Fisher's exact test), Q value = 0.55

Table S11. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	8	19	28	2	202
subtype1	1	0	0	0	13
subtype2	1	7	11	1	82
subtype3	0	4	9	0	29
subtype4	4	3	4	0	42
subtype5	2	5	4	1	36

Figure S10. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #10: 'RACE'

'Copy Number Ratio CNMF subtypes' versus 'ETHNICITY'

P value = 0.225 (Fisher's exact test), Q value = 0.55

Table S12. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	23	163
subtype1	2	9
subtype2	7	63
subtype3	2	28
subtype4	9	31
subtype5	3	32

Figure S11. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #11: 'ETHNICITY'

'Copy Number Ratio CNMF subtypes' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.324 (Kruskal-Wallis (anova)), Q value = 0.64

Table S13. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	266	73.2 (21.6)
subtype1	14	66.3 (15.6)
subtype2	101	71.5 (18.3)
subtype3	48	75.5 (29.4)
subtype4	53	76.3 (16.9)
subtype5	50	73.2 (24.6)

Figure S12. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'Copy Number Ratio CNMF subtypes' versus 'TUMOR_STATUS'

P value = 0.949 (Fisher's exact test), Q value = 0.99

Table S14. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	193	76
subtype1	11	3
subtype2	71	31
subtype3	34	14
subtype4	42	14
subtype5	35	14

Figure S13. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

'Copy Number Ratio CNMF subtypes' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 0.0298 (Fisher's exact test), Q value = 0.2

Table S15. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	18	129	115	1	24
subtype1	2	6	7	0	0
subtype2	2	40	56	1	9
subtype3	4	25	22	0	5
subtype4	8	30	16	0	5
subtype5	2	28	14	0	5

Figure S14. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'Copy Number Ratio CNMF subtypes' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.401 (Kruskal-Wallis (anova)), Q value = 0.7

Table S16. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	43	1999.7 (13.6)
subtype1	2	1991.0 (12.7)
subtype2	18	2003.5 (10.6)
subtype3	9	2000.2 (10.3)
subtype4	7	2000.3 (14.4)
subtype5	7	1991.1 (21.3)

Figure S15. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'Copy Number Ratio CNMF subtypes' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.662 (Kruskal-Wallis (anova)), Q value = 0.84

Table S17. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	90	17.5 (14.3)
subtype1	6	14.2 (10.2)
subtype2	32	16.7 (12.2)
subtype3	20	22.4 (17.9)
subtype4	14	16.0 (10.8)
subtype5	18	15.7 (16.6)

Figure S16. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'Copy Number Ratio CNMF subtypes' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.338 (Kruskal-Wallis (anova)), Q value = 0.65

Table S18. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	255	1.8 (1.1)
subtype1	13	1.8 (1.1)
subtype2	92	1.9 (1.2)
subtype3	50	1.9 (1.1)
subtype4	55	1.6 (1.0)
subtype5	45	1.9 (1.1)

Figure S17. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'Copy Number Ratio CNMF subtypes' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.527 (Kruskal-Wallis (anova)), Q value = 0.79

Table S19. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	84	21.2 (7.7)
subtype1	6	21.3 (3.3)
subtype2	31	21.6 (9.4)
subtype3	18	22.8 (8.0)
subtype4	14	18.3 (5.2)
subtype5	15	21.2 (6.5)

Figure S18. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'Copy Number Ratio CNMF subtypes' versus 'RADIATION_THERAPY_STATUS'

P value = 1 (Fisher's exact test), Q value = 1

Table S20. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	25	3
subtype1	1	0
subtype2	10	2
subtype3	4	1
subtype4	5	0
subtype5	5	0

Figure S19. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'Copy Number Ratio CNMF subtypes' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.669 (Kruskal-Wallis (anova)), Q value = 0.85

Table S21. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	255	3.6 (2.5)
subtype1	14	3.9 (2.7)
subtype2	93	3.8 (2.9)
subtype3	51	3.8 (2.5)
subtype4	51	3.0 (1.9)
subtype5	46	3.5 (2.3)

Figure S20. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'Copy Number Ratio CNMF subtypes' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.0563 (Kruskal-Wallis (anova)), Q value = 0.28

Table S22. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	106	0.1 (0.4)
subtype1	6	0.0 (0.0)
subtype2	35	0.0 (0.0)
subtype3	19	0.3 (0.7)
subtype4	25	0.1 (0.3)
subtype5	21	0.0 (0.0)

Figure S21. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'Copy Number Ratio CNMF subtypes' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.951 (Kruskal-Wallis (anova)), Q value = 0.99

Table S23. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	140	0.6 (1.0)
subtype1	8	0.4 (0.7)
subtype2	47	0.5 (0.9)
subtype3	28	0.6 (1.2)
subtype4	30	0.6 (1.0)
subtype5	27	0.5 (0.8)

Figure S22. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'Copy Number Ratio CNMF subtypes' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.26 (Kruskal-Wallis (anova)), Q value = 0.57

Table S24. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	251	2.8 (2.0)
subtype1	14	3.4 (2.6)
subtype2	96	3.0 (2.1)
subtype3	47	3.1 (2.2)
subtype4	47	2.3 (1.4)
subtype5	47	2.5 (1.7)

Figure S23. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'Copy Number Ratio CNMF subtypes' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.854 (Kruskal-Wallis (anova)), Q value = 0.93

Table S25. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	116	0.8 (1.6)
subtype1	6	0.3 (0.5)
subtype2	39	1.0 (2.4)
subtype3	20	0.7 (1.1)
subtype4	28	0.7 (1.1)
subtype5	23	0.9 (1.2)

Figure S24. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'Copy Number Ratio CNMF subtypes' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.6 (Kruskal-Wallis (anova)), Q value = 0.84

Table S26. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	110	0.1 (0.4)
subtype1	6	0.0 (0.0)
subtype2	37	0.1 (0.3)
subtype3	20	0.2 (0.5)
subtype4	25	0.0 (0.2)
subtype5	22	0.1 (0.4)

Figure S25. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'Copy Number Ratio CNMF subtypes' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.177 (Fisher's exact test), Q value = 0.49

Table S27. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	2	7	36	1	10
subtype1	1	0	1	0	0
subtype2	0	4	11	0	1
subtype3	0	1	8	0	3
subtype4	1	2	9	0	1
subtype5	0	0	7	1	5

Figure S26. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'Copy Number Ratio CNMF subtypes' versus 'MENOPAUSE_STATUS'

P value = 0.241 (Fisher's exact test), Q value = 0.55

Table S28. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	25	82	121
subtype1	0	0	3	10
subtype2	1	11	24	46
subtype3	0	4	25	17
subtype4	1	5	19	28
subtype5	1	5	11	20

Figure S27. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'Copy Number Ratio CNMF subtypes' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.172 (Fisher's exact test), Q value = 0.49

Table S29. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	71	74
subtype1	2	3
subtype2	25	23
subtype3	12	15
subtype4	23	14
subtype5	9	19

Figure S28. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'Copy Number Ratio CNMF subtypes' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.048 (Kruskal-Wallis (anova)), Q value = 0.25

Table S30. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	150	0.9 (2.1)
subtype1	6	0.3 (0.8)
subtype2	48	0.8 (1.3)
subtype3	23	1.7 (2.3)
subtype4	41	0.9 (2.8)
subtype5	32	0.8 (1.8)

Figure S29. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'Copy Number Ratio CNMF subtypes' versus 'LYMPH_NODES_EXAMINED'

P value = 0.742 (Kruskal-Wallis (anova)), Q value = 0.88

Table S31. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	170	22.3 (12.8)
subtype1	8	26.0 (13.0)
subtype2	56	21.3 (12.4)
subtype3	25	23.3 (15.1)
subtype4	43	23.5 (12.9)
subtype5	38	21.1 (11.9)

Figure S30. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'Copy Number Ratio CNMF subtypes' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.524 (Fisher's exact test), Q value = 0.79

Table S32. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	49	117
subtype1	5	6
subtype2	16	50
subtype3	12	21
subtype4	8	18
subtype5	8	22

Figure S31. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'Copy Number Ratio CNMF subtypes' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.63 (Kruskal-Wallis (anova)), Q value = 0.84

Table S33. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	293	2008.5 (4.6)
subtype1	15	2009.7 (3.9)
subtype2	111	2008.3 (4.6)
subtype3	55	2009.1 (4.2)
subtype4	59	2008.6 (4.8)
subtype5	53	2007.9 (5.2)

Figure S32. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'Copy Number Ratio CNMF subtypes' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.776 (Fisher's exact test), Q value = 0.89

Table S34. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	15	54	85
subtype1	0	4	4
subtype2	5	18	33
subtype3	1	10	17
subtype4	6	12	20
subtype5	3	10	11

Figure S33. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'Copy Number Ratio CNMF subtypes' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.461 (Kruskal-Wallis (anova)), Q value = 0.74

Table S35. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	253	161.1 (6.9)
subtype1	14	158.7 (10.6)
subtype2	97	161.3 (6.7)
subtype3	46	160.7 (6.2)
subtype4	50	160.3 (6.4)
subtype5	46	162.8 (7.3)

Figure S34. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'Copy Number Ratio CNMF subtypes' versus 'CORPUS_INVOLVEMENT'

P value = 0.95 (Fisher's exact test), Q value = 0.99

Table S36. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	92	18
subtype1	3	1
subtype2	30	6
subtype3	15	2
subtype4	25	5
subtype5	19	4

Figure S35. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'Copy Number Ratio CNMF subtypes' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.259 (Fisher's exact test), Q value = 0.57

Table S37. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	5	100	2
subtype1	0	8	0
subtype2	1	45	1
subtype3	1	21	0
subtype4	3	16	0
subtype5	0	10	1

Figure S36. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'Copy Number Ratio CNMF subtypes' versus 'CERVIX_SUV_RESULTS'

P value = 0.768 (Kruskal-Wallis (anova)), Q value = 0.89

Table S38. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

	nPatients	Mean (Std.Dev)
ALL	17	13.2 (7.2)
subtype1	1	18.5 (NA)
subtype2	7	13.9 (7.9)
subtype3	4	12.1 (3.3)
subtype4	3	15.6 (11.8)
subtype5	2	6.8 (0.4)

Figure S37. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

'Copy Number Ratio CNMF subtypes' versus 'AGE_AT_DIAGNOSIS'

P value = 0.0112 (Kruskal-Wallis (anova)), Q value = 0.13

Table S39. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	295	48.2 (13.8)
subtype1	15	45.8 (12.3)
subtype2	112	45.8 (14.0)
subtype3	56	53.4 (12.4)
subtype4	59	48.0 (12.8)
subtype5	53	48.8 (15.1)

Figure S38. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'Copy Number Ratio CNMF subtypes' versus 'CLINICAL_STAGE'

P value = 0.0246 (Fisher's exact test), Q value = 0.2

Table S40. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA1	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	5	1	1	1	34	77	37	4	8	5	7	44	1	3	42	7	12
subtype1	0	0	0	0	1	2	2	0	2	0	0	5	0	1	2	0	0
subtype2	2	1	0	0	16	23	14	1	3	4	2	22	0	1	15	2	3
subtype3	1	0	0	1	3	14	3	0	1	0	1	7	1	1	15	4	2
subtype4	1	0	0	0	5	17	13	1	2	0	3	7	0	0	6	0	4
subtype5	1	0	1	0	9	21	5	2	0	1	1	3	0	0	4	1	3

Figure S39. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

Clustering Approach #2: 'METHLYATION CNMF'

Table S41. Description of clustering approach #2: 'METHLYATION CNMF'

Cluster Labels	1	2	3
Number of samples	72	122	113

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.967 (logrank test), Q value = 0.99

Table S42. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	294	72	0.1 - 210.7 (23.2)
subtype1	70	17	0.1 - 147.4 (20.6)
subtype2	117	30	0.1 - 210.7 (20.9)
subtype3	107	25	0.4 - 177.0 (24.8)

Figure S40. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'YEARS_TO_BIRTH'

P value = 0.174 (Kruskal-Wallis (anova)), Q value = 0.49

Table S43. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	305	48.2 (13.8)
subtype1	72	47.9 (12.1)
subtype2	121	50.1 (14.5)
subtype3	112	46.4 (13.9)

Figure S41. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'METHLYATION CNMF' versus 'PATHOLOGY_T_STAGE'

P value = 0.422 (Fisher's exact test), Q value = 0.72

Table S44. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	141	72	21	10
subtype1	39	18	4	1
subtype2	47	24	12	5
subtype3	55	30	5	4

Figure S42. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_N_STAGE'

P value = 0.232 (Fisher's exact test), Q value = 0.55

Table S45. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	135	60
subtype1	36	13
subtype2	41	26
subtype3	58	21

Figure S43. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_M_STAGE'

P value = 0.36 (Fisher's exact test), Q value = 0.66

Table S46. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	116	10
subtype1	24	4
subtype2	37	2
subtype3	55	4

Figure S44. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'METHLYATION CNMF' versus 'RADIATION_THERAPY'

P value = 0.773 (Fisher's exact test), Q value = 0.89

Table S47. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	55	129
subtype1	16	31
subtype2	22	53
subtype3	17	45

Figure S45. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'RADIATION_THERAPY'

'METHLYATION CNMF' versus 'HISTOLOGICAL_TYPE'

P value = 1e-05 (Fisher's exact test), Q value = 0.00043

Table S48. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	6	254	6	21	3	17
subtype1	3	28	6	19	2	14
subtype2	0	122	0	0	0	0
subtype3	3	104	0	2	1	3

Figure S46. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'METHLYATION CNMF' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.112 (Kruskal-Wallis (anova)), Q value = 0.4

Table S49. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	19	16.4 (15.4)
subtype2	43	20.9 (15.8)
subtype3	31	13.0 (9.1)

Figure S47. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'METHLYATION CNMF' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.0366 (Kruskal-Wallis (anova)), Q value = 0.21

Table S50. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	42	1.0 (2.8)
subtype2	54	1.4 (2.8)
subtype3	63	0.7 (1.7)

Figure S48. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'METHLYATION CNMF' versus 'RACE'

P value = 0.273 (Fisher's exact test), Q value = 0.59

Table S51. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	8	20	30	2	211
subtype1	1	7	4	0	52
subtype2	4	4	11	2	85
subtype3	3	9	15	0	74

Figure S49. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #10: 'RACE'

'METHLYATION CNMF' versus 'ETHNICITY'

P value = 0.701 (Fisher's exact test), Q value = 0.86

Table S52. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	24	171
subtype1	7	37
subtype2	9	68
subtype3	8	66

Figure S50. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #11: 'ETHNICITY'

'METHLYATION CNMF' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.244 (Kruskal-Wallis (anova)), Q value = 0.55

Table S53. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	278	73.0 (21.5)
subtype1	65	76.2 (22.7)
subtype2	116	73.0 (20.1)
subtype3	97	70.9 (22.4)

Figure S51. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'METHLYATION CNMF' versus 'TUMOR_STATUS'

P value = 0.561 (Fisher's exact test), Q value = 0.8

Table S54. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	201	80
subtype1	46	21
subtype2	76	33
subtype3	79	26

Figure S52. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #13: 'TUMOR_STATUS'

'METHLYATION CNMF' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 0.375 (Fisher's exact test), Q value = 0.68

Table S55. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	18	136	120	1	24
subtype1	7	27	31	0	6
subtype2	6	62	40	0	10
subtype3	5	47	49	1	8

Figure S53. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'METHLYATION CNMF' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.497 (Kruskal-Wallis (anova)), Q value = 0.76

Table S56. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	43	1999.7 (13.6)
subtype1	9	1997.8 (11.8)
subtype2	21	1998.3 (16.4)
subtype3	13	2003.3 (9.5)

Figure S54. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'METHLYATION CNMF' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.112 (Kruskal-Wallis (anova)), Q value = 0.4

Table S57. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	19	16.4 (15.4)
subtype2	43	20.9 (15.8)
subtype3	31	13.0 (9.1)

Figure S55. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'METHLYATION CNMF' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.0617 (Kruskal-Wallis (anova)), Q value = 0.29

Table S58. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	263	1.8 (1.1)
subtype1	64	1.6 (1.1)
subtype2	99	2.0 (1.2)
subtype3	100	1.8 (1.2)

Figure S56. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'METHLYATION CNMF' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.715 (Kruskal-Wallis (anova)), Q value = 0.86

Table S59. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	85	21.1 (7.7)
subtype1	17	21.4 (6.4)
subtype2	39	21.4 (9.1)
subtype3	29	20.6 (6.5)

Figure S57. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'METHLYATION CNMF' versus 'RADIATION_THERAPY_STATUS'

P value = 1 (Fisher's exact test), Q value = 1

Table S60. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	29	3
subtype1	7	0
subtype2	8	1
subtype3	14	2

Figure S58. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'METHLYATION CNMF' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.0168 (Kruskal-Wallis (anova)), Q value = 0.16

Table S61. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	267	3.6 (2.6)
subtype1	64	3.0 (2.0)
subtype2	103	4.2 (3.0)
subtype3	100	3.4 (2.3)

Figure S59. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'METHLYATION CNMF' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.793 (Kruskal-Wallis (anova)), Q value = 0.9

Table S62. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	112	0.1 (0.3)
subtype1	26	0.0 (0.2)
subtype2	44	0.0 (0.2)
subtype3	42	0.1 (0.5)

Figure S60. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'METHLYATION CNMF' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.629 (Kruskal-Wallis (anova)), Q value = 0.84

Table S63. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	147	0.5 (0.9)
subtype1	30	0.6 (0.9)
subtype2	60	0.6 (1.2)
subtype3	57	0.4 (0.6)

Figure S61. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'METHLYATION CNMF' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.0333 (Kruskal-Wallis (anova)), Q value = 0.2

Table S64. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	262	2.8 (2.0)
subtype1	61	2.4 (1.9)
subtype2	102	3.2 (2.2)
subtype3	99	2.7 (1.9)

Figure S62. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'METHLYATION CNMF' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.76 (Kruskal-Wallis (anova)), Q value = 0.89

Table S65. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	122	0.9 (1.8)
subtype1	28	0.5 (0.7)
subtype2	48	0.9 (2.2)
subtype3	46	1.0 (1.8)

Figure S63. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'METHLYATION CNMF' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.291 (Kruskal-Wallis (anova)), Q value = 0.61

Table S66. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	116	0.1 (0.3)
subtype1	25	0.1 (0.3)
subtype2	47	0.1 (0.2)
subtype3	44	0.2 (0.4)

Figure S64. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'METHLYATION CNMF' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.366 (Fisher's exact test), Q value = 0.67

Table S67. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	2	8	40	1	12
subtype1	2	2	10	0	2
subtype2	0	3	13	1	7
subtype3	0	3	17	0	3

Figure S65. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'METHLYATION CNMF' versus 'MENOPAUSE_STATUS'

P value = 0.0814 (Fisher's exact test), Q value = 0.33

Table S68. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	25	84	125
subtype1	0	10	20	30
subtype2	2	10	37	39
subtype3	1	5	27	56

Figure S66. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'METHLYATION CNMF' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.224 (Fisher's exact test), Q value = 0.55

Table S69. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	72	80
subtype1	20	22
subtype2	20	32
subtype3	32	26

Figure S67. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'METHLYATION CNMF' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.0366 (Kruskal-Wallis (anova)), Q value = 0.21

Table S70. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	42	1.0 (2.8)
subtype2	54	1.4 (2.8)
subtype3	63	0.7 (1.7)

Figure S68. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'METHLYATION CNMF' versus 'LYMPH_NODES_EXAMINED'

P value = 0.804 (Kruskal-Wallis (anova)), Q value = 0.91

Table S71. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	180	22.3 (12.6)
subtype1	46	22.2 (10.2)
subtype2	62	22.8 (14.1)
subtype3	72	21.9 (12.7)

Figure S69. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'METHLYATION CNMF' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.0271 (Fisher's exact test), Q value = 0.2

Table S72. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	55	120
subtype1	1	17
subtype2	31	55
subtype3	23	48

Figure S70. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'METHLYATION CNMF' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.947 (Kruskal-Wallis (anova)), Q value = 0.99

Table S73. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	305	2008.3 (4.8)
subtype1	71	2008.7 (4.1)
subtype2	121	2008.2 (4.9)
subtype3	113	2008.2 (5.0)

Figure S71. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'METHLYATION CNMF' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.235 (Fisher's exact test), Q value = 0.55

Table S74. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	15	54	90
subtype1	7	17	19
subtype2	3	20	37
subtype3	5	17	34

Figure S72. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'METHLYATION CNMF' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.573 (Kruskal-Wallis (anova)), Q value = 0.81

Table S75. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	264	161.0 (7.3)
subtype1	62	161.4 (7.1)
subtype2	110	161.4 (7.0)
subtype3	92	160.1 (7.8)

Figure S73. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'METHLYATION CNMF' versus 'CORPUS_INVOLVEMENT'

P value = 0.17 (Fisher's exact test), Q value = 0.49

Table S76. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	99	19
subtype1	33	4
subtype2	32	4
subtype3	34	11

Figure S74. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'METHLYATION CNMF' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.628 (Fisher's exact test), Q value = 0.84

Table S77. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	7	104	2
subtype1	2	20	1
subtype2	3	40	0
subtype3	2	44	1

Figure S75. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'METHLYATION CNMF' versus 'CERVIX_SUV_RESULTS'

P value = 0.385 (Kruskal-Wallis (anova)), Q value = 0.68

Table S78. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

	nPatients	Mean (Std.Dev)
ALL	17	13.2 (7.2)
subtype1	2	10.8 (5.2)
subtype2	7	15.3 (7.2)
subtype3	8	12.0 (7.9)

Figure S76. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

'METHLYATION CNMF' versus 'AGE_AT_DIAGNOSIS'

P value = 0.184 (Kruskal-Wallis (anova)), Q value = 0.5

Table S79. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	307	48.3 (13.8)
subtype1	72	47.9 (12.1)
subtype2	122	50.1 (14.5)
subtype3	113	46.5 (13.9)

Figure S77. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'METHLYATION CNMF' versus 'CLINICAL_STAGE'

P value = 0.0591 (Fisher's exact test), Q value = 0.29

Table S80. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA1	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	5	1	1	1	38	78	39	5	9	5	7	44	1	3	42	9	12
subtype1	1	0	0	0	5	28	12	1	3	1	0	9	0	0	8	0	4
subtype2	4	0	1	1	12	23	13	3	2	0	4	17	1	2	22	6	5
subtype3	0	1	0	0	21	27	14	1	4	4	3	18	0	1	12	3	3

Figure S78. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #39: 'CLINICAL_STAGE'

Clustering Approach #3: 'RPPA CNMF subtypes'

Table S81. Description of clustering approach #3: 'RPPA CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	60	45	68

'RPPA CNMF subtypes' versus 'Time to Death'

P value = 0.0599 (logrank test), Q value = 0.29

Table S82. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	164	31	0.1 - 210.7 (22.8)
subtype1	57	14	0.1 - 144.2 (18.0)
subtype2	41	6	0.4 - 173.3 (27.2)
subtype3	66	11	0.1 - 210.7 (24.3)

Figure S79. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.0786 (Kruskal-Wallis (anova)), Q value = 0.32

Table S83. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	171	47.5 (13.5)
subtype1	60	44.4 (12.5)
subtype2	44	49.4 (14.0)
subtype3	67	49.1 (13.8)

Figure S80. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RPPA CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.891 (Fisher's exact test), Q value = 0.96

Table S84. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	99	32	7	3
subtype1	34	10	1	1
subtype2	26	7	3	1
subtype3	39	15	3	1

Figure S81. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.661 (Fisher's exact test), Q value = 0.84

Table S85. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	93	38
subtype1	33	11
subtype2	22	8
subtype3	38	19

Figure S82. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.55 (Fisher's exact test), Q value = 0.79

Table S86. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	84	4
subtype1	28	2
subtype2	25	0
subtype3	31	2

Figure S83. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'RPPA CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.611 (Fisher's exact test), Q value = 0.84

Table S87. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	26	60
subtype1	7	22
subtype2	7	17
subtype3	12	21

Figure S84. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

'RPPA CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.311 (Fisher's exact test), Q value = 0.63

Table S88. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	4	144	3	16	2	4
subtype1	3	51	1	3	0	2
subtype2	0	41	0	3	0	1
subtype3	1	52	2	10	2	1

Figure S85. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'RPPA CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.198 (Kruskal-Wallis (anova)), Q value = 0.51

Table S89. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	45	17.7 (13.7)
subtype1	12	12.8 (8.3)
subtype2	12	14.2 (11.0)
subtype3	21	22.5 (16.1)

Figure S86. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'RPPA CNMF subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.632 (Kruskal-Wallis (anova)), Q value = 0.84

Table S90. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	105	1.1 (2.7)
subtype1	33	0.9 (2.3)
subtype2	23	1.2 (2.1)
subtype3	49	1.3 (3.2)

Figure S87. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'RPPA CNMF subtypes' versus 'RACE'

P value = 0.307 (Fisher's exact test), Q value = 0.63

Table S91. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	3	15	12	1	125
subtype1	1	4	7	1	42
subtype2	1	7	1	0	32
subtype3	1	4	4	0	51

Figure S88. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'RACE'

'RPPA CNMF subtypes' versus 'ETHNICITY'

P value = 0.432 (Fisher's exact test), Q value = 0.72

Table S92. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	13	113
subtype1	3	43
subtype2	5	27
subtype3	5	43

Figure S89. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #11: 'ETHNICITY'

'RPPA CNMF subtypes' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.0556 (Kruskal-Wallis (anova)), Q value = 0.28

Table S93. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	158	74.7 (23.1)
subtype1	52	68.7 (17.3)
subtype2	42	75.1 (17.5)
subtype3	64	79.2 (29.0)

Figure S90. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'RPPA CNMF subtypes' versus 'TUMOR_STATUS'

P value = 0.674 (Fisher's exact test), Q value = 0.85

Table S94. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	119	36
subtype1	40	14
subtype2	32	7
subtype3	47	15

Figure S91. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

'RPPA CNMF subtypes' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 0.34 (Fisher's exact test), Q value = 0.65

Table S95. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	10	75	74	1	8
subtype1	3	26	25	1	4
subtype2	2	23	14	0	3
subtype3	5	26	35	0	1

Figure S92. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'RPPA CNMF subtypes' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.838 (Kruskal-Wallis (anova)), Q value = 0.93

Table S96. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	31	2001.5 (10.7)
subtype1	9	2003.9 (7.6)
subtype2	9	2001.0 (10.4)
subtype3	13	2000.3 (13.1)

Figure S93. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'RPPA CNMF subtypes' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.198 (Kruskal-Wallis (anova)), Q value = 0.51

Table S97. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	45	17.7 (13.7)
subtype1	12	12.8 (8.3)
subtype2	12	14.2 (11.0)
subtype3	21	22.5 (16.1)

Figure S94. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'RPPA CNMF subtypes' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.802 (Kruskal-Wallis (anova)), Q value = 0.91

Table S98. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	145	1.9 (1.2)
subtype1	49	1.9 (1.3)
subtype2	40	1.9 (1.3)
subtype3	56	1.9 (1.2)

Figure S95. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'RPPA CNMF subtypes' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.747 (Kruskal-Wallis (anova)), Q value = 0.88

Table S99. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	44	21.5 (7.9)
subtype1	11	21.9 (8.2)
subtype2	11	20.8 (8.1)
subtype3	22	21.7 (8.0)

Figure S96. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'RPPA CNMF subtypes' versus 'RADIATION_THERAPY_STATUS'

P value = 0.725 (Fisher's exact test), Q value = 0.87

Table S100. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	21	2
subtype1	9	1
subtype2	5	1
subtype3	7	0

Figure S97. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'RPPA CNMF subtypes' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.472 (Kruskal-Wallis (anova)), Q value = 0.74

Table S101. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	151	3.6 (2.6)
subtype1	48	3.4 (2.3)
subtype2	44	4.3 (3.4)
subtype3	59	3.3 (2.0)

Figure S98. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'RPPA CNMF subtypes' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.458 (Kruskal-Wallis (anova)), Q value = 0.74

Table S102. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	72	0.0 (0.2)
subtype1	26	0.1 (0.3)
subtype2	17	0.0 (0.0)
subtype3	29	0.0 (0.2)

Figure S99. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'RPPA CNMF subtypes' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.806 (Kruskal-Wallis (anova)), Q value = 0.91

Table S103. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	89	0.4 (0.8)
subtype1	29	0.3 (0.5)
subtype2	24	0.5 (1.1)
subtype3	36	0.5 (0.8)

Figure S100. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'RPPA CNMF subtypes' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.124 (Kruskal-Wallis (anova)), Q value = 0.41

Table S104. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	151	2.6 (1.9)
subtype1	50	2.4 (2.1)
subtype2	43	3.2 (2.2)
subtype3	58	2.4 (1.3)

Figure S101. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'RPPA CNMF subtypes' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.786 (Kruskal-Wallis (anova)), Q value = 0.9

Table S105. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	81	1.1 (2.2)
subtype1	30	0.9 (1.3)
subtype2	19	1.8 (3.6)
subtype3	32	0.9 (1.6)

Figure S102. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'RPPA CNMF subtypes' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.919 (Kruskal-Wallis (anova)), Q value = 0.98

Table S106. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	74	0.1 (0.4)
subtype1	26	0.1 (0.3)
subtype2	18	0.1 (0.3)
subtype3	30	0.1 (0.4)

Figure S103. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'RPPA CNMF subtypes' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.366 (Fisher's exact test), Q value = 0.67

Table S107. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	2	7	25	1	7
subtype1	2	1	5	1	2
subtype2	0	1	7	0	1
subtype3	0	5	13	0	4

Figure S104. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'RPPA CNMF subtypes' versus 'MENOPAUSE_STATUS'

P value = 0.128 (Fisher's exact test), Q value = 0.42

Table S108. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	13	46	79
subtype1	0	2	13	35
subtype2	1	6	12	17
subtype3	2	5	21	27

Figure S105. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'RPPA CNMF subtypes' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.0301 (Fisher's exact test), Q value = 0.2

Table S109. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	52	60
subtype1	22	15
subtype2	15	13
subtype3	15	32

Figure S106. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'RPPA CNMF subtypes' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.632 (Kruskal-Wallis (anova)), Q value = 0.84

Table S110. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	105	1.1 (2.7)
subtype1	33	0.9 (2.3)
subtype2	23	1.2 (2.1)
subtype3	49	1.3 (3.2)

Figure S107. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'RPPA CNMF subtypes' versus 'LYMPH_NODES_EXAMINED'

P value = 0.00935 (Kruskal-Wallis (anova)), Q value = 0.11

Table S111. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	122	20.8 (11.9)
subtype1	41	20.7 (11.3)
subtype2	25	14.8 (8.1)
subtype3	56	23.5 (12.8)

Figure S108. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'RPPA CNMF subtypes' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.0795 (Fisher's exact test), Q value = 0.32

Table S112. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	39	76
subtype1	15	29
subtype2	6	25
subtype3	18	22

Figure S109. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'RPPA CNMF subtypes' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.524 (Kruskal-Wallis (anova)), Q value = 0.79

Table S113. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	171	2008.3 (5.0)
subtype1	59	2008.5 (5.1)
subtype2	45	2008.5 (4.7)
subtype3	67	2007.8 (5.1)

Figure S110. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'RPPA CNMF subtypes' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.697 (Fisher's exact test), Q value = 0.86

Table S114. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	5	32	51
subtype1	2	10	16
subtype2	0	8	16
subtype3	3	14	19

Figure S111. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'RPPA CNMF subtypes' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.126 (Kruskal-Wallis (anova)), Q value = 0.41

Table S115. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	153	161.7 (7.1)
subtype1	51	162.4 (6.6)
subtype2	43	160.2 (8.0)
subtype3	59	162.1 (6.8)

Figure S112. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'RPPA CNMF subtypes' versus 'CORPUS_INVOLVEMENT'

P value = 0.127 (Fisher's exact test), Q value = 0.41

Table S116. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	84	12
subtype1	27	3
subtype2	26	1
subtype3	31	8

Figure S113. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'RPPA CNMF subtypes' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.509 (Fisher's exact test), Q value = 0.77

Table S117. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	3	48	1
subtype1	0	18	1
subtype2	1	15	0
subtype3	2	15	0

Figure S114. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'RPPA CNMF subtypes' versus 'CERVIX_SUV_RESULTS'

P value = 0.0662 (Kruskal-Wallis (anova)), Q value = 0.29

Table S118. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

	nPatients	Mean (Std.Dev)
ALL	11	13.2 (8.7)
subtype1	4	16.1 (7.3)
subtype2	2	23.6 (7.3)
subtype3	5	6.6 (4.4)

Figure S115. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

'RPPA CNMF subtypes' versus 'AGE_AT_DIAGNOSIS'

P value = 0.0597 (Kruskal-Wallis (anova)), Q value = 0.29

Table S119. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	173	47.7 (13.5)
subtype1	60	44.4 (12.5)
subtype2	45	49.7 (14.0)
subtype3	68	49.2 (13.7)

Figure S116. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'RPPA CNMF subtypes' versus 'CLINICAL_STAGE'

P value = 0.645 (Fisher's exact test), Q value = 0.84

Table S120. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA1	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	4	1	1	1	27	48	28	4	5	4	4	16	1	1	19	4	2
subtype1	1	1	1	0	8	15	13	0	2	1	1	8	0	0	6	1	1
subtype2	2	0	0	1	8	12	4	3	1	0	0	4	0	1	7	2	0
subtype3	1	0	0	0	11	21	11	1	2	3	3	4	1	0	6	1	1

Figure S117. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

Clustering Approach #4: 'RPPA cHierClus subtypes'

Table S121. Description of clustering approach #4: 'RPPA cHierClus subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	32	49	51	23	18

'RPPA cHierClus subtypes' versus 'Time to Death'

P value = 0.183 (logrank test), Q value = 0.5

Table S122. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	164	31	0.1 - 210.7 (22.8)
subtype1	31	6	0.4 - 144.2 (18.0)
subtype2	48	7	0.1 - 210.7 (22.6)
subtype3	48	8	0.4 - 209.6 (26.7)
subtype4	21	7	0.9 - 78.7 (24.6)
subtype5	16	3	0.1 - 147.4 (25.1)

Figure S118. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.0879 (Kruskal-Wallis (anova)), Q value = 0.35

Table S123. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	171	47.5 (13.5)
subtype1	31	44.2 (14.4)
subtype2	48	47.5 (13.5)
subtype3	51	48.9 (13.4)
subtype4	23	44.2 (12.3)
subtype5	18	53.8 (12.1)

Figure S119. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.645 (Fisher's exact test), Q value = 0.84

Table S124. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	99	32	7	3
subtype1	19	4	0	0
subtype2	31	9	2	0
subtype3	24	11	4	2
subtype4	13	3	1	1
subtype5	12	5	0	0

Figure S120. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.213 (Fisher's exact test), Q value = 0.54

Table S125. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	93	38
subtype1	17	4
subtype2	30	12
subtype3	19	14
subtype4	13	6
subtype5	14	2

Figure S121. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 1 (Fisher's exact test), Q value = 1

Table S126. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	84	4
subtype1	12	1
subtype2	24	1
subtype3	26	1
subtype4	12	1
subtype5	10	0

Figure S122. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'RPPA cHierClus subtypes' versus 'RADIATION_THERAPY'

P value = 0.303 (Fisher's exact test), Q value = 0.62

Table S127. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	26	60
subtype1	7	10
subtype2	9	16
subtype3	5	25
subtype4	3	6
subtype5	2	3

Figure S123. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

'RPPA cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.00394 (Fisher's exact test), Q value = 0.073

Table S128. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	4	144	3	16	2	4
subtype1	2	26	1	2	0	1
subtype2	2	35	1	8	2	1
subtype3	0	49	0	1	0	1
subtype4	0	23	0	0	0	0
subtype5	0	11	1	5	0	1

Figure S124. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'RPPA cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.54 (Kruskal-Wallis (anova)), Q value = 0.79

Table S129. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	45	17.7 (13.7)
subtype1	8	10.9 (7.4)
subtype2	15	17.3 (10.7)
subtype3	14	20.8 (16.6)
subtype4	6	20.6 (20.0)
subtype5	2	18.0 (11.3)

Figure S125. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'RPPA cHierClus subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.177 (Kruskal-Wallis (anova)), Q value = 0.49

Table S130. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	105	1.1 (2.7)
subtype1	16	1.3 (2.9)
subtype2	39	1.3 (3.5)
subtype3	25	1.2 (1.6)
subtype4	14	0.6 (1.6)
subtype5	11	0.7 (2.4)

Figure S126. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'RPPA cHierClus subtypes' versus 'RACE'

P value = 0.209 (Fisher's exact test), Q value = 0.54

Table S131. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	3	15	12	1	125
subtype1	0	4	3	1	23
subtype2	2	1	2	0	37
subtype3	1	4	3	0	37
subtype4	0	2	4	0	15
subtype5	0	4	0	0	13

Figure S127. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'RACE'

'RPPA cHierClus subtypes' versus 'ETHNICITY'

P value = 0.473 (Fisher's exact test), Q value = 0.74

Table S132. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	13	113
subtype1	1	28
subtype2	3	29
subtype3	6	29
subtype4	1	14
subtype5	2	13

Figure S128. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #11: 'ETHNICITY'

'RPPA cHierClus subtypes' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.542 (Kruskal-Wallis (anova)), Q value = 0.79

Table S133. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	158	74.7 (23.1)
subtype1	27	68.4 (19.0)
subtype2	45	76.0 (20.0)
subtype3	48	76.6 (27.3)
subtype4	20	72.6 (12.8)
subtype5	18	77.8 (31.8)

Figure S129. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'RPPA cHierClus subtypes' versus 'TUMOR_STATUS'

P value = 0.53 (Fisher's exact test), Q value = 0.79

Table S134. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	119	36
subtype1	24	6
subtype2	36	9
subtype3	36	9
subtype4	13	7
subtype5	10	5

Figure S130. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

'RPPA cHierClus subtypes' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 0.551 (Fisher's exact test), Q value = 0.79

Table S135. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	10	75	74	1	8
subtype1	2	15	12	0	3
subtype2	5	16	26	0	2
subtype3	2	26	17	0	2
subtype4	1	9	10	1	1
subtype5	0	9	9	0	0

Figure S131. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'RPPA cHierClus subtypes' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.706 (Kruskal-Wallis (anova)), Q value = 0.86

Table S136. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	31	2001.5 (10.7)
subtype1	6	2002.2 (8.0)
subtype2	11	2000.3 (14.3)
subtype3	10	2004.3 (8.2)
subtype4	1	2000.0 (NA)
subtype5	3	1996.3 (11.8)

Figure S132. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'RPPA cHierClus subtypes' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.54 (Kruskal-Wallis (anova)), Q value = 0.79

Table S137. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	45	17.7 (13.7)
subtype1	8	10.9 (7.4)
subtype2	15	17.3 (10.7)
subtype3	14	20.8 (16.6)
subtype4	6	20.6 (20.0)
subtype5	2	18.0 (11.3)

Figure S133. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'RPPA cHierClus subtypes' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.471 (Kruskal-Wallis (anova)), Q value = 0.74

Table S138. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	145	1.9 (1.2)
subtype1	27	1.6 (1.0)
subtype2	39	2.0 (1.3)
subtype3	44	2.0 (1.3)
subtype4	19	1.8 (1.2)
subtype5	16	1.6 (1.1)

Figure S134. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'RPPA cHierClus subtypes' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.808 (Kruskal-Wallis (anova)), Q value = 0.91

Table S139. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	44	21.5 (7.9)
subtype1	7	23.4 (9.6)
subtype2	14	20.9 (7.1)
subtype3	15	20.5 (7.9)
subtype4	5	21.4 (5.9)
subtype5	3	25.3 (13.6)

Figure S135. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'RPPA cHierClus subtypes' versus 'RADIATION_THERAPY_STATUS'

P value = 1 (Fisher's exact test), Q value = 1

Table S140. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	21	2
subtype1	5	1
subtype2	6	0
subtype3	7	1
subtype4	3	0
subtype5	0	0

Figure S136. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'RPPA cHierClus subtypes' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.622 (Kruskal-Wallis (anova)), Q value = 0.84

Table S141. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	151	3.6 (2.6)
subtype1	25	3.3 (2.1)
subtype2	41	3.3 (1.7)
subtype3	49	3.9 (2.9)
subtype4	21	3.3 (2.9)
subtype5	15	4.8 (3.6)

Figure S137. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'RPPA cHierClus subtypes' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.224 (Kruskal-Wallis (anova)), Q value = 0.55

Table S142. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	72	0.0 (0.2)
subtype1	13	0.2 (0.4)
subtype2	21	0.0 (0.2)
subtype3	22	0.0 (0.0)
subtype4	11	0.0 (0.0)
subtype5	5	0.0 (0.0)

Figure S138. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'RPPA cHierClus subtypes' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.809 (Kruskal-Wallis (anova)), Q value = 0.91

Table S143. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	89	0.4 (0.8)
subtype1	15	0.3 (0.6)
subtype2	25	0.4 (0.9)
subtype3	28	0.6 (1.1)
subtype4	14	0.4 (0.5)
subtype5	7	0.4 (0.5)

Figure S139. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'RPPA cHierClus subtypes' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.114 (Kruskal-Wallis (anova)), Q value = 0.4

Table S144. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	151	2.6 (1.9)
subtype1	26	2.3 (1.7)
subtype2	40	2.4 (1.4)
subtype3	48	2.9 (1.8)
subtype4	21	2.4 (2.6)
subtype5	16	3.3 (2.3)

Figure S140. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'RPPA cHierClus subtypes' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.655 (Kruskal-Wallis (anova)), Q value = 0.84

Table S145. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	81	1.1 (2.2)
subtype1	16	1.1 (1.5)
subtype2	23	0.7 (1.2)
subtype3	23	1.2 (2.5)
subtype4	12	0.9 (1.2)
subtype5	7	2.6 (4.7)

Figure S141. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'RPPA cHierClus subtypes' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.638 (Kruskal-Wallis (anova)), Q value = 0.84

Table S146. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	74	0.1 (0.4)
subtype1	12	0.0 (0.0)
subtype2	23	0.1 (0.3)
subtype3	22	0.1 (0.5)
subtype4	12	0.2 (0.4)
subtype5	5	0.2 (0.4)

Figure S142. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'RPPA cHierClus subtypes' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.251 (Fisher's exact test), Q value = 0.56

Table S147. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	2	7	25	1	7
subtype1	2	1	3	0	0
subtype2	0	4	10	0	3
subtype3	0	2	7	0	2
subtype4	0	0	3	1	2
subtype5	0	0	2	0	0

Figure S143. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'RPPA cHierClus subtypes' versus 'MENOPAUSE_STATUS'

P value = 0.873 (Fisher's exact test), Q value = 0.95

Table S148. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	13	46	79
subtype1	0	2	7	19
subtype2	1	3	15	23
subtype3	2	6	13	18
subtype4	0	1	5	12
subtype5	0	1	6	7

Figure S144. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'RPPA cHierClus subtypes' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.425 (Fisher's exact test), Q value = 0.72

Table S149. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	52	60
subtype1	13	8
subtype2	13	22
subtype3	12	17
subtype4	6	6
subtype5	8	7

Figure S145. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'RPPA cHierClus subtypes' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.177 (Kruskal-Wallis (anova)), Q value = 0.49

Table S150. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	105	1.1 (2.7)
subtype1	16	1.3 (2.9)
subtype2	39	1.3 (3.5)
subtype3	25	1.2 (1.6)
subtype4	14	0.6 (1.6)
subtype5	11	0.7 (2.4)

Figure S146. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'RPPA cHierClus subtypes' versus 'LYMPH_NODES_EXAMINED'

P value = 0.236 (Kruskal-Wallis (anova)), Q value = 0.55

Table S151. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	122	20.8 (11.9)
subtype1	22	18.4 (9.4)
subtype2	45	24.2 (12.6)
subtype3	26	19.3 (10.9)
subtype4	16	19.6 (14.9)
subtype5	13	17.3 (8.9)

Figure S147. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'RPPA cHierClus subtypes' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.0376 (Fisher's exact test), Q value = 0.21

Table S152. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	39	76
subtype1	7	18
subtype2	16	14
subtype3	10	25
subtype4	6	11
subtype5	0	8

Figure S148. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'RPPA cHierClus subtypes' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.841 (Kruskal-Wallis (anova)), Q value = 0.93

Table S153. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	171	2008.3 (5.0)
subtype1	32	2008.3 (5.5)
subtype2	47	2008.6 (4.4)
subtype3	51	2008.1 (5.2)
subtype4	23	2007.1 (5.6)
subtype5	18	2009.2 (3.8)

Figure S149. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'RPPA cHierClus subtypes' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.552 (Fisher's exact test), Q value = 0.79

Table S154. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	5	32	51
subtype1	0	4	8
subtype2	3	12	13
subtype3	1	6	18
subtype4	0	6	8
subtype5	1	4	4

Figure S150. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'RPPA cHierClus subtypes' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.167 (Kruskal-Wallis (anova)), Q value = 0.49

Table S155. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	153	161.7 (7.1)
subtype1	26	162.4 (6.5)
subtype2	43	162.3 (5.7)
subtype3	46	160.1 (8.7)
subtype4	20	164.4 (6.1)
subtype5	18	160.3 (7.2)

Figure S151. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'RPPA cHierClus subtypes' versus 'CORPUS_INVOLVEMENT'

P value = 0.455 (Fisher's exact test), Q value = 0.74

Table S156. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	84	12
subtype1	16	1
subtype2	24	3
subtype3	25	3
subtype4	9	1
subtype5	10	4

Figure S152. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'RPPA cHierClus subtypes' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.00359 (Fisher's exact test), Q value = 0.073

Table S157. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	3	48	1
subtype1	0	9	1
subtype2	1	10	0
subtype3	0	22	0
subtype4	0	6	0
subtype5	2	1	0

Figure S153. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'RPPA cHierClus subtypes' versus 'AGE_AT_DIAGNOSIS'

P value = 0.117 (Kruskal-Wallis (anova)), Q value = 0.4

Table S158. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	173	47.7 (13.5)
subtype1	32	44.8 (14.5)
subtype2	49	47.7 (13.5)
subtype3	51	48.9 (13.4)
subtype4	23	44.2 (12.3)
subtype5	18	53.8 (12.1)

Figure S154. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'RPPA cHierClus subtypes' versus 'CLINICAL_STAGE'

P value = 0.13 (Fisher's exact test), Q value = 0.42

Table S159. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA1	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	4	1	1	1	27	48	28	4	5	4	4	16	1	1	19	4	2
subtype1	1	1	1	0	4	7	9	0	1	0	0	3	0	0	3	0	1
subtype2	1	0	0	0	8	20	8	0	1	1	2	3	0	0	3	0	1
subtype3	1	0	0	1	8	10	4	2	1	0	0	8	1	1	10	3	0
subtype4	1	0	0	0	5	5	3	0	1	1	1	2	0	0	3	1	0
subtype5	0	0	0	0	2	6	4	2	1	2	1	0	0	0	0	0	0

Figure S155. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

Clustering Approach #5: 'RNAseq CNMF subtypes'

Table S160. Description of clustering approach #5: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3	4
Number of samples	61	103	69	71

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.44 (logrank test), Q value = 0.72

Table S161. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	291	72	0.1 - 210.7 (23.5)
subtype1	58	12	0.1 - 146.9 (21.8)
subtype2	98	24	0.2 - 210.7 (26.7)
subtype3	68	15	0.4 - 147.4 (25.0)
subtype4	67	21	0.1 - 154.3 (18.7)

Figure S156. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 8.51e-06 (Kruskal-Wallis (anova)), Q value = 0.00043

Table S162. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	302	48.2 (13.9)
subtype1	61	53.2 (12.6)
subtype2	102	50.7 (14.9)
subtype3	69	46.2 (12.1)
subtype4	70	41.9 (12.4)

Figure S157. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.661 (Fisher's exact test), Q value = 0.84

Table S163. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	140	71	20	10
subtype1	24	17	5	3
subtype2	47	25	8	4
subtype3	40	14	2	1
subtype4	29	15	5	2

Figure S158. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.377 (Fisher's exact test), Q value = 0.68

Table S164. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	133	60
subtype1	26	13
subtype2	44	26
subtype3	35	10
subtype4	28	11

Figure S159. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.267 (Fisher's exact test), Q value = 0.58

Table S165. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	116	10
subtype1	25	4
subtype2	43	1
subtype3	23	2
subtype4	25	3

Figure S160. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'RNAseq CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.888 (Fisher's exact test), Q value = 0.96

Table S166. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	55	127
subtype1	9	26
subtype2	16	39
subtype3	15	30
subtype4	15	32

Figure S161. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

'RNAseq CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 1e-05 (Fisher's exact test), Q value = 0.00043

Table S167. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	5	252	6	21	3	17
subtype1	3	50	2	3	1	2
subtype2	0	103	0	0	0	0
subtype3	2	28	4	18	2	15
subtype4	0	71	0	0	0	0

Figure S162. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'RNAseq CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.106 (Kruskal-Wallis (anova)), Q value = 0.39

Table S168. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	19	22.4 (17.0)
subtype2	37	19.7 (15.0)
subtype3	17	12.3 (10.6)
subtype4	20	12.6 (9.4)

Figure S163. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'RNAseq CNMF subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.0146 (Kruskal-Wallis (anova)), Q value = 0.15

Table S169. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	32	1.2 (2.9)
subtype2	57	1.5 (2.3)
subtype3	44	0.8 (2.7)
subtype4	26	0.3 (0.6)

Figure S164. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'RNAseq CNMF subtypes' versus 'RACE'

P value = 0.00901 (Fisher's exact test), Q value = 0.11

Table S170. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	7	20	30	2	209
subtype1	2	7	9	0	38
subtype2	1	7	10	0	69
subtype3	0	2	1	0	55
subtype4	4	4	10	2	47

Figure S165. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'RACE'

'RNAseq CNMF subtypes' versus 'ETHNICITY'

P value = 0.255 (Fisher's exact test), Q value = 0.57

Table S171. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	24	169
subtype1	2	39
subtype2	8	59
subtype3	5	32
subtype4	9	39

Figure S166. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #11: 'ETHNICITY'

'RNAseq CNMF subtypes' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.694 (Kruskal-Wallis (anova)), Q value = 0.86

Table S172. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	275	72.9 (21.5)
subtype1	55	72.3 (24.8)
subtype2	93	73.3 (25.0)
subtype3	63	73.3 (15.8)
subtype4	64	72.4 (18.1)

Figure S167. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'RNAseq CNMF subtypes' versus 'TUMOR_STATUS'

P value = 0.38 (Fisher's exact test), Q value = 0.68

Table S173. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	198	80
subtype1	36	21
subtype2	69	22
subtype3	48	17
subtype4	45	20

Figure S168. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

'RNAseq CNMF subtypes' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 0.00049 (Fisher's exact test), Q value = 0.014

Table S174. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	18	135	118	1	24
subtype1	2	22	32	1	2
subtype2	7	62	26	0	6
subtype3	7	30	25	0	7
subtype4	2	21	35	0	9

Figure S169. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'RNAseq CNMF subtypes' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.836 (Kruskal-Wallis (anova)), Q value = 0.93

Table S175. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	43	1999.7 (13.6)
subtype1	8	1999.2 (11.7)
subtype2	17	1997.5 (16.6)
subtype3	11	2001.5 (12.6)
subtype4	7	2002.6 (10.6)

Figure S170. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'RNAseq CNMF subtypes' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.106 (Kruskal-Wallis (anova)), Q value = 0.39

Table S176. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	19	22.4 (17.0)
subtype2	37	19.7 (15.0)
subtype3	17	12.3 (10.6)
subtype4	20	12.6 (9.4)

Figure S171. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'RNAseq CNMF subtypes' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.255 (Kruskal-Wallis (anova)), Q value = 0.57

Table S177. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	261	1.8 (1.1)
subtype1	50	2.0 (1.2)
subtype2	92	1.9 (1.2)
subtype3	60	1.7 (1.2)
subtype4	59	1.7 (1.1)

Figure S172. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'RNAseq CNMF subtypes' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.226 (Kruskal-Wallis (anova)), Q value = 0.55

Table S178. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	85	21.1 (7.7)
subtype1	17	23.3 (7.5)
subtype2	36	21.4 (8.3)
subtype3	15	19.9 (4.4)
subtype4	17	19.4 (8.7)

Figure S173. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'RNAseq CNMF subtypes' versus 'RADIATION_THERAPY_STATUS'

P value = 0.733 (Fisher's exact test), Q value = 0.87

Table S179. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	29	3
subtype1	7	0
subtype2	10	1
subtype3	4	0
subtype4	8	2

Figure S174. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'RNAseq CNMF subtypes' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.0261 (Kruskal-Wallis (anova)), Q value = 0.2

Table S180. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	264	3.6 (2.6)
subtype1	51	3.4 (2.4)
subtype2	91	4.1 (2.7)
subtype3	61	3.0 (2.1)
subtype4	61	3.7 (2.8)

Figure S175. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'RNAseq CNMF subtypes' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.164 (Kruskal-Wallis (anova)), Q value = 0.49

Table S181. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	112	0.1 (0.3)
subtype1	19	0.0 (0.0)
subtype2	46	0.2 (0.5)
subtype3	26	0.0 (0.2)
subtype4	21	0.0 (0.0)

Figure S176. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'RNAseq CNMF subtypes' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.772 (Kruskal-Wallis (anova)), Q value = 0.89

Table S182. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	147	0.5 (0.9)
subtype1	25	0.5 (1.1)
subtype2	62	0.5 (1.0)
subtype3	28	0.5 (0.8)
subtype4	32	0.6 (0.9)

Figure S177. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'RNAseq CNMF subtypes' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.0381 (Kruskal-Wallis (anova)), Q value = 0.21

Table S183. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	260	2.9 (2.1)
subtype1	52	2.6 (1.7)
subtype2	89	3.1 (1.9)
subtype3	59	2.4 (2.1)
subtype4	60	3.1 (2.4)

Figure S178. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'RNAseq CNMF subtypes' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.815 (Kruskal-Wallis (anova)), Q value = 0.91

Table S184. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	122	0.9 (1.8)
subtype1	21	0.8 (1.4)
subtype2	51	1.2 (2.5)
subtype3	27	0.6 (0.8)
subtype4	23	0.5 (0.9)

Figure S179. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'RNAseq CNMF subtypes' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.192 (Kruskal-Wallis (anova)), Q value = 0.51

Table S185. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	116	0.1 (0.3)
subtype1	20	0.2 (0.4)
subtype2	49	0.1 (0.4)
subtype3	26	0.1 (0.3)
subtype4	21	0.0 (0.0)

Figure S180. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'RNAseq CNMF subtypes' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.359 (Fisher's exact test), Q value = 0.66

Table S186. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	2	8	40	1	12
subtype1	1	1	5	0	4
subtype2	0	2	16	1	5
subtype3	1	2	10	0	3
subtype4	0	3	9	0	0

Figure S181. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'RNAseq CNMF subtypes' versus 'MENOPAUSE_STATUS'

P value = 0.0481 (Fisher's exact test), Q value = 0.25

Table S187. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	25	82	124
subtype1	0	6	22	23
subtype2	2	8	34	37
subtype3	1	7	19	30
subtype4	0	4	7	34

Figure S182. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'RNAseq CNMF subtypes' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.819 (Fisher's exact test), Q value = 0.92

Table S188. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	71	79
subtype1	15	19
subtype2	25	29
subtype3	18	21
subtype4	13	10

Figure S183. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'RNAseq CNMF subtypes' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.0146 (Kruskal-Wallis (anova)), Q value = 0.15

Table S189. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	32	1.2 (2.9)
subtype2	57	1.5 (2.3)
subtype3	44	0.8 (2.7)
subtype4	26	0.3 (0.6)

Figure S184. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'RNAseq CNMF subtypes' versus 'LYMPH_NODES_EXAMINED'

P value = 0.174 (Kruskal-Wallis (anova)), Q value = 0.49

Table S190. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	178	22.4 (12.6)
subtype1	37	19.5 (12.6)
subtype2	60	23.0 (14.4)
subtype3	49	24.0 (11.1)
subtype4	32	22.0 (10.9)

Figure S185. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'RNAseq CNMF subtypes' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.0321 (Fisher's exact test), Q value = 0.2

Table S191. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	55	119
subtype1	15	22
subtype2	21	53
subtype3	1	16
subtype4	18	28

Figure S186. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'RNAseq CNMF subtypes' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.347 (Kruskal-Wallis (anova)), Q value = 0.66

Table S192. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	302	2008.3 (4.8)
subtype1	61	2009.1 (4.0)
subtype2	102	2008.0 (5.4)
subtype3	68	2008.7 (4.1)
subtype4	71	2007.6 (4.9)

Figure S187. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'RNAseq CNMF subtypes' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.0953 (Fisher's exact test), Q value = 0.36

Table S193. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	15	53	89
subtype1	0	13	20
subtype2	3	15	31
subtype3	8	15	18
subtype4	4	10	20

Figure S188. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'RNAseq CNMF subtypes' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.291 (Kruskal-Wallis (anova)), Q value = 0.61

Table S194. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	261	161.0 (7.1)
subtype1	52	159.7 (6.7)
subtype2	88	160.5 (7.3)
subtype3	61	162.2 (7.3)
subtype4	60	161.7 (6.9)

Figure S189. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'RNAseq CNMF subtypes' versus 'CORPUS_INVOLVEMENT'

P value = 0.0691 (Fisher's exact test), Q value = 0.3

Table S195. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	99	19
subtype1	19	6
subtype2	32	9
subtype3	28	4
subtype4	20	0

Figure S190. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'RNAseq CNMF subtypes' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.396 (Fisher's exact test), Q value = 0.7

Table S196. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	7	102	2
subtype1	3	16	0
subtype2	2	34	0
subtype3	1	20	1
subtype4	1	32	1

Figure S191. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'RNAseq CNMF subtypes' versus 'CERVIX_SUV_RESULTS'

P value = 0.231 (Kruskal-Wallis (anova)), Q value = 0.55

Table S197. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

	nPatients	Mean (Std.Dev)
ALL	17	13.2 (7.2)
subtype1	3	13.4 (1.3)
subtype2	9	11.7 (8.4)
subtype3	1	7.1 (NA)
subtype4	4	18.0 (6.0)

Figure S192. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

'RNAseq CNMF subtypes' versus 'AGE_AT_DIAGNOSIS'

P value = 1.34e-05 (Kruskal-Wallis (anova)), Q value = 0.00052

Table S198. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	304	48.2 (13.8)
subtype1	61	53.2 (12.6)
subtype2	103	50.8 (14.9)
subtype3	69	46.2 (12.1)
subtype4	71	42.2 (12.5)

Figure S193. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'RNAseq CNMF subtypes' versus 'CLINICAL_STAGE'

P value = 0.0304 (Fisher's exact test), Q value = 0.2

Table S199. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA1	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	5	1	1	1	38	77	39	5	9	5	7	43	1	2	42	9	12
subtype1	1	1	1	0	5	13	9	1	1	3	4	7	0	1	8	2	3
subtype2	1	0	0	1	15	22	9	4	2	1	2	12	1	0	20	4	5
subtype3	1	0	0	0	5	31	10	0	2	0	0	10	0	0	5	1	3
subtype4	2	0	0	0	13	11	11	0	4	1	1	14	0	1	9	2	1

Figure S194. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

Clustering Approach #6: 'RNAseq cHierClus subtypes'

Table S200. Description of clustering approach #6: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	72	195	37

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.0302 (logrank test), Q value = 0.2

Table S201. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	291	72	0.1 - 210.7 (23.5)
subtype1	70	17	0.5 - 137.2 (20.6)
subtype2	185	42	0.1 - 210.7 (24.6)
subtype3	36	13	0.4 - 99.9 (17.2)

Figure S195. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.00802 (Kruskal-Wallis (anova)), Q value = 0.11

Table S202. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	302	48.2 (13.9)
subtype1	72	47.6 (11.6)
subtype2	193	49.7 (14.5)
subtype3	37	41.4 (12.7)

Figure S196. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.704 (Fisher's exact test), Q value = 0.86

Table S203. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	140	71	20	10
subtype1	39	20	3	2
subtype2	86	42	16	8
subtype3	15	9	1	0

Figure S197. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.349 (Fisher's exact test), Q value = 0.66

Table S204. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	133	60
subtype1	38	13
subtype2	78	42
subtype3	17	5

Figure S198. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.0524 (Fisher's exact test), Q value = 0.27

Table S205. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	116	10
subtype1	27	5
subtype2	77	3
subtype3	12	2

Figure S199. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'RNAseq cHierClus subtypes' versus 'RADIATION_THERAPY'

P value = 1 (Fisher's exact test), Q value = 1

Table S206. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	55	127
subtype1	14	33
subtype2	35	79
subtype3	6	15

Figure S200. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

'RNAseq cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 1e-05 (Fisher's exact test), Q value = 0.00043

Table S207. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	5	252	6	21	3	17
subtype1	5	22	5	21	3	16
subtype2	0	193	1	0	0	1
subtype3	0	37	0	0	0	0

Figure S201. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'RNAseq cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.437 (Kruskal-Wallis (anova)), Q value = 0.72

Table S208. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	18	14.2 (12.0)
subtype2	63	19.0 (15.2)
subtype3	12	13.8 (9.9)

Figure S202. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'RNAseq cHierClus subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.0308 (Kruskal-Wallis (anova)), Q value = 0.2

Table S209. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	44	0.7 (2.2)
subtype2	97	1.4 (2.6)
subtype3	18	0.2 (0.5)

Figure S203. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'RNAseq cHierClus subtypes' versus 'RACE'

P value = 0.307 (Fisher's exact test), Q value = 0.63

Table S210. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	7	20	30	2	209
subtype1	1	6	3	0	53
subtype2	4	13	20	2	130
subtype3	2	1	7	0	26

Figure S204. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'RACE'

'RNAseq cHierClus subtypes' versus 'ETHNICITY'

P value = 0.896 (Fisher's exact test), Q value = 0.96

Table S211. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	24	169
subtype1	6	38
subtype2	16	109
subtype3	2	22

Figure S205. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #11: 'ETHNICITY'

'RNAseq cHierClus subtypes' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.677 (Kruskal-Wallis (anova)), Q value = 0.85

Table S212. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	275	72.9 (21.5)
subtype1	66	72.5 (17.0)
subtype2	177	72.9 (23.6)
subtype3	32	73.6 (17.6)

Figure S206. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'RNAseq cHierClus subtypes' versus 'TUMOR_STATUS'

P value = 0.425 (Fisher's exact test), Q value = 0.72

Table S213. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	198	80
subtype1	45	22
subtype2	130	46
subtype3	23	12

Figure S207. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

'RNAseq cHierClus subtypes' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 0.157 (Fisher's exact test), Q value = 0.47

Table S214. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	18	135	118	1	24
subtype1	7	30	29	0	6
subtype2	10	93	69	0	16
subtype3	1	12	20	1	2

Figure S208. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'RNAseq cHierClus subtypes' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.569 (Kruskal-Wallis (anova)), Q value = 0.81

Table S215. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	43	1999.7 (13.6)
subtype1	9	2002.1 (12.6)
subtype2	30	1999.1 (14.8)
subtype3	4	1999.0 (5.2)

Figure S209. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'RNAseq cHierClus subtypes' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.437 (Kruskal-Wallis (anova)), Q value = 0.72

Table S216. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	18	14.2 (12.0)
subtype2	63	19.0 (15.2)
subtype3	12	13.8 (9.9)

Figure S210. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'RNAseq cHierClus subtypes' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.0944 (Kruskal-Wallis (anova)), Q value = 0.36

Table S217. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	261	1.8 (1.1)
subtype1	65	1.6 (1.1)
subtype2	166	1.9 (1.2)
subtype3	30	1.9 (1.2)

Figure S211. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'RNAseq cHierClus subtypes' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.655 (Kruskal-Wallis (anova)), Q value = 0.84

Table S218. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	85	21.1 (7.7)
subtype1	17	20.8 (5.7)
subtype2	56	21.5 (8.1)
subtype3	12	19.8 (8.3)

Figure S212. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'RNAseq cHierClus subtypes' versus 'RADIATION_THERAPY_STATUS'

P value = 1 (Fisher's exact test), Q value = 1

Table S219. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	29	3
subtype1	3	0
subtype2	20	2
subtype3	6	1

Figure S213. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'RNAseq cHierClus subtypes' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.0475 (Kruskal-Wallis (anova)), Q value = 0.25

Table S220. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	264	3.6 (2.6)
subtype1	65	3.0 (2.0)
subtype2	167	3.8 (2.7)
subtype3	32	4.0 (2.8)

Figure S214. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'RNAseq cHierClus subtypes' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.574 (Kruskal-Wallis (anova)), Q value = 0.81

Table S221. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	112	0.1 (0.3)
subtype1	25	0.0 (0.2)
subtype2	74	0.1 (0.4)
subtype3	13	0.0 (0.0)

Figure S215. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'RNAseq cHierClus subtypes' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.13 (Kruskal-Wallis (anova)), Q value = 0.42

Table S222. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	147	0.5 (0.9)
subtype1	29	0.6 (0.9)
subtype2	98	0.5 (0.9)
subtype3	20	0.8 (1.1)

Figure S216. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'RNAseq cHierClus subtypes' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.0188 (Kruskal-Wallis (anova)), Q value = 0.17

Table S223. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	260	2.9 (2.1)
subtype1	63	2.3 (1.8)
subtype2	165	3.0 (2.0)
subtype3	32	3.2 (2.4)

Figure S217. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'RNAseq cHierClus subtypes' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.616 (Kruskal-Wallis (anova)), Q value = 0.84

Table S224. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	122	0.9 (1.8)
subtype1	27	0.7 (1.0)
subtype2	80	0.9 (2.1)
subtype3	15	0.6 (1.0)

Figure S218. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'RNAseq cHierClus subtypes' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.855 (Kruskal-Wallis (anova)), Q value = 0.93

Table S225. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	116	0.1 (0.3)
subtype1	24	0.1 (0.3)
subtype2	79	0.1 (0.4)
subtype3	13	0.1 (0.3)

Figure S219. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'RNAseq cHierClus subtypes' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.426 (Fisher's exact test), Q value = 0.72

Table S226. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	2	8	40	1	12
subtype1	2	1	9	0	2
subtype2	0	5	26	1	9
subtype3	0	2	5	0	1

Figure S220. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'RNAseq cHierClus subtypes' versus 'MENOPAUSE_STATUS'

P value = 0.0234 (Fisher's exact test), Q value = 0.2

Table S227. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	25	82	124
subtype1	0	8	19	34
subtype2	2	15	61	72
subtype3	1	2	2	18

Figure S221. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'RNAseq cHierClus subtypes' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.639 (Fisher's exact test), Q value = 0.84

Table S228. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	71	79
subtype1	23	21
subtype2	39	50
subtype3	9	8

Figure S222. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'RNAseq cHierClus subtypes' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.0308 (Kruskal-Wallis (anova)), Q value = 0.2

Table S229. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	44	0.7 (2.2)
subtype2	97	1.4 (2.6)
subtype3	18	0.2 (0.5)

Figure S223. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'RNAseq cHierClus subtypes' versus 'LYMPH_NODES_EXAMINED'

P value = 0.342 (Kruskal-Wallis (anova)), Q value = 0.65

Table S230. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	178	22.4 (12.6)
subtype1	48	21.2 (10.2)
subtype2	109	22.2 (13.4)
subtype3	21	25.7 (13.0)

Figure S224. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'RNAseq cHierClus subtypes' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.00374 (Fisher's exact test), Q value = 0.073

Table S231. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	55	119
subtype1	0	17
subtype2	47	89
subtype3	8	13

Figure S225. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'RNAseq cHierClus subtypes' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.0135 (Kruskal-Wallis (anova)), Q value = 0.14

Table S232. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	302	2008.3 (4.8)
subtype1	71	2009.5 (3.5)
subtype2	194	2008.2 (4.8)
subtype3	37	2006.4 (5.8)

Figure S226. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'RNAseq cHierClus subtypes' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.0336 (Fisher's exact test), Q value = 0.2

Table S233. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	15	53	89
subtype1	7	17	21
subtype2	4	32	59
subtype3	4	4	9

Figure S227. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'RNAseq cHierClus subtypes' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.146 (Kruskal-Wallis (anova)), Q value = 0.46

Table S234. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	261	161.0 (7.1)
subtype1	64	161.1 (6.9)
subtype2	167	160.5 (7.2)
subtype3	30	163.5 (6.8)

Figure S228. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'RNAseq cHierClus subtypes' versus 'CORPUS_INVOLVEMENT'

P value = 0.342 (Fisher's exact test), Q value = 0.65

Table S235. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	99	19
subtype1	31	6
subtype2	56	13
subtype3	12	0

Figure S229. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'RNAseq cHierClus subtypes' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.117 (Fisher's exact test), Q value = 0.4

Table S236. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	7	102	2
subtype1	3	22	1
subtype2	4	67	0
subtype3	0	13	1

Figure S230. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'RNAseq cHierClus subtypes' versus 'CERVIX_SUV_RESULTS'

P value = 0.226 (Kruskal-Wallis (anova)), Q value = 0.55

Table S237. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

	nPatients	Mean (Std.Dev)
ALL	17	13.2 (7.2)
subtype1	1	7.1 (NA)
subtype2	13	12.5 (7.0)
subtype3	3	18.4 (7.3)

Figure S231. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

'RNAseq cHierClus subtypes' versus 'AGE_AT_DIAGNOSIS'

P value = 0.00655 (Kruskal-Wallis (anova)), Q value = 0.1

Table S238. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	304	48.2 (13.8)
subtype1	72	47.6 (11.6)
subtype2	195	49.8 (14.4)
subtype3	37	41.4 (12.7)

Figure S232. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'RNAseq cHierClus subtypes' versus 'CLINICAL_STAGE'

P value = 0.106 (Fisher's exact test), Q value = 0.39

Table S239. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA1	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	5	1	1	1	38	77	39	5	9	5	7	43	1	2	42	9	12
subtype1	1	0	0	0	6	28	10	1	3	1	2	8	0	0	6	1	5
subtype2	4	1	1	1	24	44	20	4	3	2	4	30	1	2	33	8	7
subtype3	0	0	0	0	8	5	9	0	3	2	1	5	0	0	3	0	0

Figure S233. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

Clustering Approach #7: 'MIRSEQ CNMF'

Table S240. Description of clustering approach #7: 'MIRSEQ CNMF'

Cluster Labels	1	2	3
Number of samples	108	78	121

'MIRSEQ CNMF' versus 'Time to Death'

P value = 0.889 (logrank test), Q value = 0.96

Table S241. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	294	72	0.1 - 210.7 (23.2)
subtype1	106	25	0.2 - 210.7 (21.8)
subtype2	75	17	0.1 - 155.8 (21.7)
subtype3	113	30	0.1 - 209.6 (24.0)

Figure S234. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CNMF' versus 'YEARS_TO_BIRTH'

P value = 0.619 (Kruskal-Wallis (anova)), Q value = 0.84

Table S242. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	305	48.2 (13.8)
subtype1	108	48.1 (11.9)
subtype2	78	49.1 (13.7)
subtype3	119	47.7 (15.5)

Figure S235. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRSEQ CNMF' versus 'PATHOLOGY_T_STAGE'

P value = 0.172 (Fisher's exact test), Q value = 0.49

Table S243. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	141	72	21	10
subtype1	53	28	10	6
subtype2	46	14	3	1
subtype3	42	30	8	3

Figure S236. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_N_STAGE'

P value = 0.609 (Fisher's exact test), Q value = 0.84

Table S244. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	135	60
subtype1	48	18
subtype2	40	22
subtype3	47	20

Figure S237. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_M_STAGE'

P value = 0.0987 (Fisher's exact test), Q value = 0.37

Table S245. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	116	10
subtype1	39	7
subtype2	36	1
subtype3	41	2

Figure S238. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'MIRSEQ CNMF' versus 'RADIATION_THERAPY'

P value = 0.955 (Fisher's exact test), Q value = 0.99

Table S246. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	55	129
subtype1	23	54
subtype2	13	28
subtype3	19	47

Figure S239. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'RADIATION_THERAPY'

'MIRSEQ CNMF' versus 'HISTOLOGICAL_TYPE'

P value = 1e-05 (Fisher's exact test), Q value = 0.00043

Table S247. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	6	254	6	21	3	17
subtype1	5	66	5	17	3	12
subtype2	1	68	1	4	0	4
subtype3	0	120	0	0	0	1

Figure S240. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'MIRSEQ CNMF' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.0659 (Kruskal-Wallis (anova)), Q value = 0.29

Table S248. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	28	14.2 (13.1)
subtype2	25	22.5 (14.9)
subtype3	40	16.4 (13.8)

Figure S241. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'MIRSEQ CNMF' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.242 (Kruskal-Wallis (anova)), Q value = 0.55

Table S249. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	55	0.8 (2.2)
subtype2	55	1.4 (2.5)
subtype3	49	0.9 (2.5)

Figure S242. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'MIRSEQ CNMF' versus 'RACE'

P value = 0.217 (Fisher's exact test), Q value = 0.55

Table S250. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	8	20	30	2	211
subtype1	1	8	10	0	73
subtype2	1	8	6	0	54
subtype3	6	4	14	2	84

Figure S243. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #10: 'RACE'

'MIRSEQ CNMF' versus 'ETHNICITY'

P value = 0.674 (Fisher's exact test), Q value = 0.85

Table S251. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	24	171
subtype1	7	57
subtype2	5	46
subtype3	12	68

Figure S244. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #11: 'ETHNICITY'

'MIRSEQ CNMF' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.551 (Kruskal-Wallis (anova)), Q value = 0.79

Table S252. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	278	73.0 (21.5)
subtype1	102	73.7 (20.1)
subtype2	68	72.6 (27.4)
subtype3	108	72.6 (18.7)

Figure S245. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'MIRSEQ CNMF' versus 'TUMOR_STATUS'

P value = 0.58 (Fisher's exact test), Q value = 0.81

Table S253. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	201	80
subtype1	69	31
subtype2	54	17
subtype3	78	32

Figure S246. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #13: 'TUMOR_STATUS'

'MIRSEQ CNMF' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 0.00417 (Fisher's exact test), Q value = 0.074

Table S254. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	18	136	120	1	24
subtype1	8	42	43	0	14
subtype2	8	41	27	1	1
subtype3	2	53	50	0	9

Figure S247. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'MIRSEQ CNMF' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.125 (Kruskal-Wallis (anova)), Q value = 0.41

Table S255. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	43	1999.7 (13.6)
subtype1	12	2002.9 (11.5)
subtype2	10	1994.6 (11.2)
subtype3	21	2000.3 (15.5)

Figure S248. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'MIRSEQ CNMF' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.0659 (Kruskal-Wallis (anova)), Q value = 0.29

Table S256. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	28	14.2 (13.1)
subtype2	25	22.5 (14.9)
subtype3	40	16.4 (13.8)

Figure S249. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'MIRSEQ CNMF' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.043 (Kruskal-Wallis (anova)), Q value = 0.23

Table S257. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	263	1.8 (1.1)
subtype1	97	1.6 (1.0)
subtype2	68	1.9 (1.2)
subtype3	98	1.9 (1.2)

Figure S250. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'MIRSEQ CNMF' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.892 (Kruskal-Wallis (anova)), Q value = 0.96

Table S258. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	85	21.1 (7.7)
subtype1	25	20.8 (6.7)
subtype2	22	20.7 (5.8)
subtype3	38	21.6 (9.2)

Figure S251. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'MIRSEQ CNMF' versus 'RADIATION_THERAPY_STATUS'

P value = 0.551 (Fisher's exact test), Q value = 0.79

Table S259. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	29	3
subtype1	7	0
subtype2	6	0
subtype3	16	3

Figure S252. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'MIRSEQ CNMF' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.506 (Kruskal-Wallis (anova)), Q value = 0.77

Table S260. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	267	3.6 (2.6)
subtype1	97	3.5 (2.9)
subtype2	70	3.6 (2.3)
subtype3	100	3.6 (2.5)

Figure S253. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'MIRSEQ CNMF' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.756 (Kruskal-Wallis (anova)), Q value = 0.89

Table S261. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	112	0.1 (0.3)
subtype1	31	0.1 (0.2)
subtype2	35	0.1 (0.5)
subtype3	46	0.1 (0.2)

Figure S254. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'MIRSEQ CNMF' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.0206 (Kruskal-Wallis (anova)), Q value = 0.19

Table S262. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	147	0.5 (0.9)
subtype1	43	0.8 (1.2)
subtype2	45	0.4 (0.8)
subtype3	59	0.4 (0.8)

Figure S255. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'MIRSEQ CNMF' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.649 (Kruskal-Wallis (anova)), Q value = 0.84

Table S263. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	262	2.8 (2.0)
subtype1	92	2.9 (2.5)
subtype2	69	2.7 (1.7)
subtype3	101	2.9 (1.9)

Figure S256. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'MIRSEQ CNMF' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.462 (Kruskal-Wallis (anova)), Q value = 0.74

Table S264. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	122	0.9 (1.8)
subtype1	34	0.8 (1.4)
subtype2	39	1.1 (2.3)
subtype3	49	0.7 (1.7)

Figure S257. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'MIRSEQ CNMF' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.338 (Kruskal-Wallis (anova)), Q value = 0.65

Table S265. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	116	0.1 (0.3)
subtype1	31	0.1 (0.2)
subtype2	37	0.2 (0.5)
subtype3	48	0.1 (0.3)

Figure S258. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'MIRSEQ CNMF' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.0901 (Fisher's exact test), Q value = 0.35

Table S266. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	2	8	40	1	12
subtype1	2	1	10	0	2
subtype2	0	1	16	1	7
subtype3	0	6	14	0	3

Figure S259. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'MIRSEQ CNMF' versus 'MENOPAUSE_STATUS'

P value = 0.727 (Fisher's exact test), Q value = 0.87

Table S267. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	25	84	125
subtype1	1	12	26	48
subtype2	1	6	26	30
subtype3	1	7	32	47

Figure S260. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'MIRSEQ CNMF' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.607 (Fisher's exact test), Q value = 0.84

Table S268. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	72	80
subtype1	25	29
subtype2	22	29
subtype3	25	22

Figure S261. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'MIRSEQ CNMF' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.242 (Kruskal-Wallis (anova)), Q value = 0.55

Table S269. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	55	0.8 (2.2)
subtype2	55	1.4 (2.5)
subtype3	49	0.9 (2.5)

Figure S262. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'MIRSEQ CNMF' versus 'LYMPH_NODES_EXAMINED'

P value = 0.847 (Kruskal-Wallis (anova)), Q value = 0.93

Table S270. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	180	22.3 (12.6)
subtype1	60	21.1 (11.6)
subtype2	61	23.6 (14.1)
subtype3	59	22.1 (11.8)

Figure S263. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'MIRSEQ CNMF' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.0335 (Fisher's exact test), Q value = 0.2

Table S271. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	55	120
subtype1	6	34
subtype2	18	32
subtype3	31	54

Figure S264. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'MIRSEQ CNMF' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.000899 (Kruskal-Wallis (anova)), Q value = 0.022

Table S272. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	305	2008.3 (4.8)
subtype1	107	2009.4 (4.5)
subtype2	77	2008.1 (4.7)
subtype3	121	2007.6 (4.9)

Figure S265. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'MIRSEQ CNMF' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.731 (Fisher's exact test), Q value = 0.87

Table S273. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	15	54	90
subtype1	8	22	36
subtype2	3	15	19
subtype3	4	17	35

Figure S266. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'MIRSEQ CNMF' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.904 (Kruskal-Wallis (anova)), Q value = 0.96

Table S274. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	264	161.0 (7.3)
subtype1	97	160.9 (7.0)
subtype2	63	161.0 (8.0)
subtype3	104	161.0 (7.2)

Figure S267. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'MIRSEQ CNMF' versus 'CORPUS_INVOLVEMENT'

P value = 0.577 (Fisher's exact test), Q value = 0.81

Table S275. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	99	19
subtype1	35	7
subtype2	32	8
subtype3	32	4

Figure S268. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'MIRSEQ CNMF' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.841 (Fisher's exact test), Q value = 0.93

Table S276. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	7	104	2
subtype1	3	39	0
subtype2	1	20	0
subtype3	3	45	2

Figure S269. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'MIRSEQ CNMF' versus 'CERVIX_SUV_RESULTS'

P value = 0.367 (Kruskal-Wallis (anova)), Q value = 0.67

Table S277. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

	nPatients	Mean (Std.Dev)
ALL	17	13.2 (7.2)
subtype1	3	16.8 (11.0)
subtype2	6	9.5 (5.7)
subtype3	8	14.7 (6.3)

Figure S270. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

'MIRSEQ CNMF' versus 'AGE_AT_DIAGNOSIS'

P value = 0.698 (Kruskal-Wallis (anova)), Q value = 0.86

Table S278. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	307	48.3 (13.8)
subtype1	108	48.1 (11.9)
subtype2	78	49.1 (13.7)
subtype3	121	47.9 (15.4)

Figure S271. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'MIRSEQ CNMF' versus 'CLINICAL_STAGE'

P value = 0.00923 (Fisher's exact test), Q value = 0.11

Table S279. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA1	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	5	1	1	1	38	78	39	5	9	5	7	44	1	3	42	9	12
subtype1	3	0	0	0	10	30	17	1	2	1	5	11	0	2	13	4	9
subtype2	0	1	1	1	8	28	11	1	3	2	1	6	0	0	10	2	1
subtype3	2	0	0	0	20	20	11	3	4	2	1	27	1	1	19	3	2

Figure S272. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #39: 'CLINICAL_STAGE'

Clustering Approach #8: 'MIRSEQ CHIERARCHICAL'

Table S280. Description of clustering approach #8: 'MIRSEQ CHIERARCHICAL'

Cluster Labels	1	2	3	4	5
Number of samples	60	102	74	26	45

'MIRSEQ CHIERARCHICAL' versus 'Time to Death'

P value = 0.692 (logrank test), Q value = 0.86

Table S281. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	294	72	0.1 - 210.7 (23.2)
subtype1	59	15	0.5 - 137.2 (18.8)
subtype2	97	21	0.1 - 160.4 (20.9)
subtype3	70	17	0.1 - 210.7 (31.3)
subtype4	24	7	0.3 - 209.6 (25.1)
subtype5	44	12	0.1 - 144.2 (17.9)

Figure S273. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CHIERARCHICAL' versus 'YEARS_TO_BIRTH'

P value = 0.000125 (Kruskal-Wallis (anova)), Q value = 0.0038

Table S282. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	305	48.2 (13.8)
subtype1	60	46.5 (12.5)
subtype2	101	50.4 (14.3)
subtype3	74	51.2 (14.1)
subtype4	25	49.9 (11.3)
subtype5	45	39.7 (11.6)

Figure S274. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_T_STAGE'

P value = 0.00365 (Fisher's exact test), Q value = 0.073

Table S283. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	141	72	21	10
subtype1	35	13	2	1
subtype2	32	29	12	6
subtype3	47	11	2	1
subtype4	8	9	2	1
subtype5	19	10	3	1

Figure S275. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_N_STAGE'

P value = 0.23 (Fisher's exact test), Q value = 0.55

Table S284. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	135	60
subtype1	32	10
subtype2	33	22
subtype3	38	20
subtype4	11	2
subtype5	21	6

Figure S276. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_M_STAGE'

P value = 0.354 (Fisher's exact test), Q value = 0.66

Table S285. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	116	10
subtype1	18	2
subtype2	29	2
subtype3	38	1
subtype4	12	2
subtype5	19	3

Figure S277. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'RADIATION_THERAPY'

P value = 0.445 (Fisher's exact test), Q value = 0.73

Table S286. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	55	129
subtype1	11	28
subtype2	17	51
subtype3	14	19
subtype4	6	11
subtype5	7	20

Figure S278. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'RADIATION_THERAPY'

'MIRSEQ CHIERARCHICAL' versus 'HISTOLOGICAL_TYPE'

P value = 1e-05 (Fisher's exact test), Q value = 0.00043

Table S287. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	6	254	6	21	3	17
subtype1	5	13	4	19	3	16
subtype2	1	99	1	1	0	0
subtype3	0	74	0	0	0	0
subtype4	0	24	1	0	0	1
subtype5	0	44	0	1	0	0

Figure S279. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'MIRSEQ CHIERARCHICAL' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.157 (Kruskal-Wallis (anova)), Q value = 0.47

Table S288. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	18	14.7 (11.6)
subtype2	36	19.9 (17.8)
subtype3	21	20.2 (10.9)
subtype4	5	7.7 (8.3)
subtype5	13	13.4 (9.5)

Figure S280. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'MIRSEQ CHIERARCHICAL' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.134 (Kruskal-Wallis (anova)), Q value = 0.42

Table S289. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	40	0.6 (1.9)
subtype2	40	1.7 (3.1)
subtype3	50	1.0 (1.9)
subtype4	8	0.4 (0.7)
subtype5	21	0.9 (3.1)

Figure S281. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'MIRSEQ CHIERARCHICAL' versus 'RACE'

P value = 0.33 (Fisher's exact test), Q value = 0.65

Table S290. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	8	20	30	2	211
subtype1	0	5	2	0	45
subtype2	3	5	13	0	66
subtype3	1	6	7	0	51
subtype4	1	2	3	1	17
subtype5	3	2	5	1	32

Figure S282. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'RACE'

'MIRSEQ CHIERARCHICAL' versus 'ETHNICITY'

P value = 0.412 (Fisher's exact test), Q value = 0.71

Table S291. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	24	171
subtype1	4	31
subtype2	10	53
subtype3	3	50
subtype4	3	16
subtype5	4	21

Figure S283. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #11: 'ETHNICITY'

'MIRSEQ CHIERARCHICAL' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.933 (Kruskal-Wallis (anova)), Q value = 0.98

Table S292. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	278	73.0 (21.5)
subtype1	54	71.7 (16.8)
subtype2	96	72.8 (20.3)
subtype3	65	75.2 (28.7)
subtype4	24	68.6 (17.0)
subtype5	39	74.6 (19.2)

Figure S284. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'MIRSEQ CHIERARCHICAL' versus 'TUMOR_STATUS'

P value = 0.713 (Fisher's exact test), Q value = 0.86

Table S293. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	201	80
subtype1	40	17
subtype2	67	26
subtype3	51	16
subtype4	14	9
subtype5	29	12

Figure S285. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #13: 'TUMOR_STATUS'

'MIRSEQ CHIERARCHICAL' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 0.0228 (Fisher's exact test), Q value = 0.2

Table S294. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	18	136	120	1	24
subtype1	6	27	21	0	6
subtype2	4	46	38	0	8
subtype3	6	39	28	0	1
subtype4	0	14	9	0	3
subtype5	2	10	24	1	6

Figure S286. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'MIRSEQ CHIERARCHICAL' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.147 (Kruskal-Wallis (anova)), Q value = 0.46

Table S295. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	43	1999.7 (13.6)
subtype1	9	2002.1 (12.6)
subtype2	16	2001.1 (17.6)
subtype3	8	1992.9 (10.7)
subtype4	3	1998.0 (10.5)
subtype5	7	2002.0 (7.7)

Figure S287. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'MIRSEQ CHIERARCHICAL' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.157 (Kruskal-Wallis (anova)), Q value = 0.47

Table S296. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	93	17.4 (14.1)
subtype1	18	14.7 (11.6)
subtype2	36	19.9 (17.8)
subtype3	21	20.2 (10.9)
subtype4	5	7.7 (8.3)
subtype5	13	13.4 (9.5)

Figure S288. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'MIRSEQ CHIERARCHICAL' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.669 (Kruskal-Wallis (anova)), Q value = 0.85

Table S297. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	263	1.8 (1.1)
subtype1	55	1.7 (1.1)
subtype2	85	1.9 (1.2)
subtype3	63	1.8 (1.1)
subtype4	22	1.6 (1.0)
subtype5	38	2.0 (1.3)

Figure S289. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'MIRSEQ CHIERARCHICAL' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.764 (Kruskal-Wallis (anova)), Q value = 0.89

Table S298. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	85	21.1 (7.7)
subtype1	17	19.8 (4.7)
subtype2	32	21.9 (8.9)
subtype3	16	21.6 (6.9)
subtype4	5	24.8 (10.1)
subtype5	15	19.5 (8.0)

Figure S290. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'MIRSEQ CHIERARCHICAL' versus 'RADIATION_THERAPY_STATUS'

P value = 0.858 (Fisher's exact test), Q value = 0.93

Table S299. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	29	3
subtype1	3	0
subtype2	7	0
subtype3	9	1
subtype4	2	0
subtype5	8	2

Figure S291. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'MIRSEQ CHIERARCHICAL' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.0673 (Kruskal-Wallis (anova)), Q value = 0.29

Table S300. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	267	3.6 (2.6)
subtype1	54	2.9 (1.9)
subtype2	88	4.2 (3.2)
subtype3	63	3.6 (2.0)
subtype4	20	3.7 (2.2)
subtype5	42	3.2 (2.6)

Figure S292. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'MIRSEQ CHIERARCHICAL' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.481 (Kruskal-Wallis (anova)), Q value = 0.74

Table S301. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	112	0.1 (0.3)
subtype1	20	0.1 (0.2)
subtype2	27	0.0 (0.2)
subtype3	39	0.2 (0.5)
subtype4	9	0.0 (0.0)
subtype5	17	0.0 (0.0)

Figure S293. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'MIRSEQ CHIERARCHICAL' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.179 (Kruskal-Wallis (anova)), Q value = 0.49

Table S302. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	147	0.5 (0.9)
subtype1	23	0.7 (0.9)
subtype2	41	0.7 (1.3)
subtype3	47	0.4 (0.7)
subtype4	12	0.3 (0.7)
subtype5	24	0.5 (0.7)

Figure S294. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'MIRSEQ CHIERARCHICAL' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.016 (Kruskal-Wallis (anova)), Q value = 0.16

Table S303. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	262	2.8 (2.0)
subtype1	52	2.2 (1.8)
subtype2	83	3.3 (2.4)
subtype3	66	2.7 (1.6)
subtype4	21	3.2 (1.9)
subtype5	40	2.6 (2.2)

Figure S295. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'MIRSEQ CHIERARCHICAL' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.936 (Kruskal-Wallis (anova)), Q value = 0.98

Table S304. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	122	0.9 (1.8)
subtype1	22	0.5 (0.7)
subtype2	31	1.5 (3.1)
subtype3	41	0.7 (1.1)
subtype4	9	0.7 (1.4)
subtype5	19	0.6 (0.9)

Figure S296. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'MIRSEQ CHIERARCHICAL' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.958 (Kruskal-Wallis (anova)), Q value = 0.99

Table S305. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	116	0.1 (0.3)
subtype1	19	0.1 (0.3)
subtype2	30	0.1 (0.3)
subtype3	41	0.1 (0.4)
subtype4	9	0.1 (0.3)
subtype5	17	0.1 (0.2)

Figure S297. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'MIRSEQ CHIERARCHICAL' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.428 (Fisher's exact test), Q value = 0.72

Table S306. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	2	8	40	1	12
subtype1	2	0	7	0	2
subtype2	0	3	8	1	4
subtype3	0	3	16	0	5
subtype4	0	0	2	0	0
subtype5	0	2	7	0	1

Figure S298. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'MIRSEQ CHIERARCHICAL' versus 'MENOPAUSE_STATUS'

P value = 0.0134 (Fisher's exact test), Q value = 0.14

Table S307. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	25	84	125
subtype1	0	5	16	30
subtype2	1	9	29	37
subtype3	1	5	30	23
subtype4	0	4	5	9
subtype5	1	2	4	26

Figure S299. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'MIRSEQ CHIERARCHICAL' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.378 (Fisher's exact test), Q value = 0.68

Table S308. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	72	80
subtype1	21	16
subtype2	17	24
subtype3	19	29
subtype4	5	3
subtype5	10	8

Figure S300. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'MIRSEQ CHIERARCHICAL' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.134 (Kruskal-Wallis (anova)), Q value = 0.42

Table S309. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	159	1.0 (2.4)
subtype1	40	0.6 (1.9)
subtype2	40	1.7 (3.1)
subtype3	50	1.0 (1.9)
subtype4	8	0.4 (0.7)
subtype5	21	0.9 (3.1)

Figure S301. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'MIRSEQ CHIERARCHICAL' versus 'LYMPH_NODES_EXAMINED'

P value = 0.319 (Kruskal-Wallis (anova)), Q value = 0.63

Table S310. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	180	22.3 (12.6)
subtype1	43	21.8 (10.6)
subtype2	48	20.6 (12.0)
subtype3	55	24.6 (14.6)
subtype4	10	15.9 (8.3)
subtype5	24	23.8 (12.7)

Figure S302. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'MIRSEQ CHIERARCHICAL' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.123 (Fisher's exact test), Q value = 0.41

Table S311. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	55	120
subtype1	0	11
subtype2	21	46
subtype3	20	34
subtype4	4	12
subtype5	10	17

Figure S303. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'MIRSEQ CHIERARCHICAL' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.0235 (Kruskal-Wallis (anova)), Q value = 0.2

Table S312. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	305	2008.3 (4.8)
subtype1	59	2009.4 (4.0)
subtype2	102	2008.9 (4.3)
subtype3	73	2007.3 (5.0)
subtype4	26	2008.0 (5.9)
subtype5	45	2007.5 (5.1)

Figure S304. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'MIRSEQ CHIERARCHICAL' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.317 (Fisher's exact test), Q value = 0.63

Table S313. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	15	54	90
subtype1	6	14	18
subtype2	4	17	33
subtype3	1	14	19
subtype4	0	5	7
subtype5	4	4	13

Figure S305. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'MIRSEQ CHIERARCHICAL' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.686 (Kruskal-Wallis (anova)), Q value = 0.85

Table S314. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	264	161.0 (7.3)
subtype1	52	161.1 (7.2)
subtype2	92	160.7 (7.6)
subtype3	62	160.5 (7.5)
subtype4	21	160.1 (7.4)
subtype5	37	162.7 (6.3)

Figure S306. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'MIRSEQ CHIERARCHICAL' versus 'CORPUS_INVOLVEMENT'

P value = 0.339 (Fisher's exact test), Q value = 0.65

Table S315. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	99	19
subtype1	27	4
subtype2	22	5
subtype3	32	8
subtype4	5	2
subtype5	13	0

Figure S307. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'MIRSEQ CHIERARCHICAL' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.458 (Fisher's exact test), Q value = 0.74

Table S316. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	7	104	2
subtype1	2	18	1
subtype2	2	36	0
subtype3	2	20	0
subtype4	1	8	0
subtype5	0	22	1

Figure S308. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'MIRSEQ CHIERARCHICAL' versus 'CERVIX_SUV_RESULTS'

P value = 0.242 (Kruskal-Wallis (anova)), Q value = 0.55

Table S317. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

	nPatients	Mean (Std.Dev)
ALL	17	13.2 (7.2)
subtype1	1	7.1 (NA)
subtype2	5	15.8 (8.6)
subtype3	8	10.5 (5.5)
subtype5	3	18.4 (7.3)

Figure S309. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

'MIRSEQ CHIERARCHICAL' versus 'AGE_AT_DIAGNOSIS'

P value = 9.35e-05 (Kruskal-Wallis (anova)), Q value = 0.003

Table S318. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	307	48.3 (13.8)
subtype1	60	46.5 (12.5)
subtype2	102	50.4 (14.3)
subtype3	74	51.2 (14.1)
subtype4	26	50.4 (11.4)
subtype5	45	39.7 (11.6)

Figure S310. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'MIRSEQ CHIERARCHICAL' versus 'CLINICAL_STAGE'

P value = 0.00075 (Fisher's exact test), Q value = 0.019

Table S319. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA1	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	5	1	1	1	38	78	39	5	9	5	7	44	1	3	42	9	12
subtype1	1	0	0	0	5	26	9	1	2	1	0	5	0	0	6	1	3
subtype2	2	0	0	0	7	17	12	3	1	0	3	18	0	3	19	5	6
subtype3	0	1	1	1	13	25	8	1	2	1	1	5	0	0	11	2	1
subtype4	2	0	0	0	6	1	2	0	1	1	2	5	1	0	3	0	2
subtype5	0	0	0	0	7	9	8	0	3	2	1	11	0	0	3	1	0

Figure S311. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #39: 'CLINICAL_STAGE'

Clustering Approach #9: 'MIRseq Mature CNMF subtypes'

Table S320. Description of clustering approach #9: 'MIRseq Mature CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	104	82	108

'MIRseq Mature CNMF subtypes' versus 'Time to Death'

P value = 0.779 (logrank test), Q value = 0.89

Table S321. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	283	69	0.1 - 210.7 (23.8)
subtype1	102	20	0.3 - 210.7 (20.4)
subtype2	78	19	0.1 - 155.8 (20.8)
subtype3	103	30	0.1 - 177.0 (28.5)

Figure S312. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.0366 (Kruskal-Wallis (anova)), Q value = 0.21

Table S322. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	293	48.2 (13.8)
subtype1	104	49.6 (13.1)
subtype2	81	50.0 (14.3)
subtype3	108	45.4 (13.8)

Figure S313. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.211 (Fisher's exact test), Q value = 0.54

Table S323. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	136	68	19	9
subtype1	44	25	10	5
subtype2	44	13	3	2
subtype3	48	30	6	2

Figure S314. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.547 (Fisher's exact test), Q value = 0.79

Table S324. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	130	57
subtype1	42	14
subtype2	42	20
subtype3	46	23

Figure S315. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.114 (Fisher's exact test), Q value = 0.4

Table S325. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	112	9
subtype1	32	6
subtype2	36	1
subtype3	44	2

Figure S316. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'MIRseq Mature CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.976 (Fisher's exact test), Q value = 1

Table S326. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	53	124
subtype1	22	51
subtype2	12	26
subtype3	19	47

Figure S317. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

'MIRseq Mature CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 1e-05 (Fisher's exact test), Q value = 0.00043

Table S327. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	6	242	6	20	3	17
subtype1	5	65	4	14	3	13
subtype2	1	69	2	6	0	4
subtype3	0	108	0	0	0	0

Figure S318. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'MIRseq Mature CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.302 (Kruskal-Wallis (anova)), Q value = 0.62

Table S328. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	88	17.2 (14.0)
subtype1	28	15.4 (13.6)
subtype2	26	19.4 (12.2)
subtype3	34	16.9 (15.7)

Figure S319. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'MIRseq Mature CNMF subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.478 (Kruskal-Wallis (anova)), Q value = 0.74

Table S329. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	155	1.0 (2.3)
subtype1	45	0.7 (2.2)
subtype2	54	1.2 (2.4)
subtype3	56	1.0 (2.5)

Figure S320. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'MIRseq Mature CNMF subtypes' versus 'RACE'

P value = 0.0823 (Fisher's exact test), Q value = 0.33

Table S330. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	7	19	27	2	205
subtype1	1	7	13	0	69
subtype2	1	8	3	0	60
subtype3	5	4	11	2	76

Figure S321. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'RACE'

'MIRseq Mature CNMF subtypes' versus 'ETHNICITY'

P value = 0.24 (Fisher's exact test), Q value = 0.55

Table S331. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	24	161
subtype1	8	53
subtype2	4	52
subtype3	12	56

Figure S322. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #11: 'ETHNICITY'

'MIRseq Mature CNMF subtypes' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.476 (Kruskal-Wallis (anova)), Q value = 0.74

Table S332. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	266	72.9 (21.4)
subtype1	96	73.7 (20.6)
subtype2	74	72.3 (27.6)
subtype3	96	72.6 (16.5)

Figure S323. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'MIRseq Mature CNMF subtypes' versus 'TUMOR_STATUS'

P value = 0.556 (Fisher's exact test), Q value = 0.8

Table S333. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	194	75
subtype1	64	30
subtype2	55	20
subtype3	75	25

Figure S324. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

'MIRseq Mature CNMF subtypes' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 0.00354 (Fisher's exact test), Q value = 0.073

Table S334. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	16	128	118	1	23
subtype1	5	37	47	0	13
subtype2	9	43	29	0	1
subtype3	2	48	42	1	9

Figure S325. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'MIRseq Mature CNMF subtypes' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.422 (Kruskal-Wallis (anova)), Q value = 0.72

Table S335. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	41	2000.0 (13.9)
subtype1	16	2001.9 (15.9)
subtype2	9	1996.2 (16.2)
subtype3	16	2000.1 (10.5)

Figure S326. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'MIRseq Mature CNMF subtypes' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.302 (Kruskal-Wallis (anova)), Q value = 0.62

Table S336. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	88	17.2 (14.0)
subtype1	28	15.4 (13.6)
subtype2	26	19.4 (12.2)
subtype3	34	16.9 (15.7)

Figure S327. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'MIRseq Mature CNMF subtypes' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.444 (Kruskal-Wallis (anova)), Q value = 0.73

Table S337. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	255	1.8 (1.1)
subtype1	89	1.8 (1.2)
subtype2	70	1.9 (1.2)
subtype3	96	1.9 (1.2)

Figure S328. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'MIRseq Mature CNMF subtypes' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.95 (Kruskal-Wallis (anova)), Q value = 0.99

Table S338. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	80	20.9 (7.6)
subtype1	26	21.0 (6.9)
subtype2	22	20.2 (6.8)
subtype3	32	21.2 (8.7)

Figure S329. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'MIRseq Mature CNMF subtypes' versus 'RADIATION_THERAPY_STATUS'

P value = 1 (Fisher's exact test), Q value = 1

Table S339. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	27	3
subtype1	5	0
subtype2	6	1
subtype3	16	2

Figure S330. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'MIRseq Mature CNMF subtypes' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.494 (Kruskal-Wallis (anova)), Q value = 0.76

Table S340. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	257	3.6 (2.5)
subtype1	90	3.4 (2.5)
subtype2	74	3.7 (2.3)
subtype3	93	3.7 (2.7)

Figure S331. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'MIRseq Mature CNMF subtypes' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.756 (Kruskal-Wallis (anova)), Q value = 0.89

Table S341. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	107	0.1 (0.3)
subtype1	24	0.0 (0.2)
subtype2	35	0.0 (0.2)
subtype3	48	0.1 (0.5)

Figure S332. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'MIRseq Mature CNMF subtypes' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.194 (Kruskal-Wallis (anova)), Q value = 0.51

Table S342. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	143	0.6 (1.0)
subtype1	37	0.8 (1.3)
subtype2	47	0.4 (0.7)
subtype3	59	0.5 (0.9)

Figure S333. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'MIRseq Mature CNMF subtypes' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.957 (Kruskal-Wallis (anova)), Q value = 0.99

Table S343. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	252	2.8 (2.0)
subtype1	87	2.8 (2.1)
subtype2	73	2.8 (1.9)
subtype3	92	2.9 (2.1)

Figure S334. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'MIRseq Mature CNMF subtypes' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.188 (Kruskal-Wallis (anova)), Q value = 0.51

Table S344. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	116	0.8 (1.8)
subtype1	27	0.7 (1.0)
subtype2	40	1.1 (2.2)
subtype3	49	0.6 (1.7)

Figure S335. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'MIRseq Mature CNMF subtypes' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.982 (Kruskal-Wallis (anova)), Q value = 1

Table S345. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	112	0.1 (0.3)
subtype1	26	0.1 (0.3)
subtype2	37	0.1 (0.4)
subtype3	49	0.1 (0.3)

Figure S336. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'MIRseq Mature CNMF subtypes' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.397 (Fisher's exact test), Q value = 0.7

Table S346. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	1	7	39	1	12
subtype1	1	1	8	0	2
subtype2	0	3	12	1	7
subtype3	0	3	19	0	3

Figure S337. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'MIRseq Mature CNMF subtypes' versus 'MENOPAUSE_STATUS'

P value = 0.012 (Fisher's exact test), Q value = 0.14

Table S347. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	25	78	120
subtype1	1	15	26	39
subtype2	0	6	29	28
subtype3	2	4	23	53

Figure S338. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'MIRseq Mature CNMF subtypes' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.695 (Fisher's exact test), Q value = 0.86

Table S348. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	69	74
subtype1	20	25
subtype2	23	26
subtype3	26	23

Figure S339. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'MIRseq Mature CNMF subtypes' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.478 (Kruskal-Wallis (anova)), Q value = 0.74

Table S349. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	155	1.0 (2.3)
subtype1	45	0.7 (2.2)
subtype2	54	1.2 (2.4)
subtype3	56	1.0 (2.5)

Figure S340. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'MIRseq Mature CNMF subtypes' versus 'LYMPH_NODES_EXAMINED'

P value = 0.642 (Kruskal-Wallis (anova)), Q value = 0.84

Table S350. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	172	22.7 (12.6)
subtype1	52	20.8 (11.4)
subtype2	60	23.5 (12.8)
subtype3	60	23.5 (13.5)

Figure S341. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'MIRseq Mature CNMF subtypes' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.00884 (Fisher's exact test), Q value = 0.11

Table S351. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	52	114
subtype1	5	34
subtype2	21	30
subtype3	26	50

Figure S342. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'MIRseq Mature CNMF subtypes' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.00872 (Kruskal-Wallis (anova)), Q value = 0.11

Table S352. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	292	2008.3 (4.8)
subtype1	103	2009.3 (4.4)
subtype2	81	2008.2 (4.7)
subtype3	108	2007.5 (5.1)

Figure S343. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'MIRseq Mature CNMF subtypes' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.625 (Fisher's exact test), Q value = 0.84

Table S353. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	15	54	85
subtype1	6	20	33
subtype2	3	18	19
subtype3	6	16	33

Figure S344. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'MIRseq Mature CNMF subtypes' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.708 (Kruskal-Wallis (anova)), Q value = 0.86

Table S354. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	253	160.8 (7.3)
subtype1	91	160.8 (6.5)
subtype2	69	160.9 (8.2)
subtype3	93	160.8 (7.4)

Figure S345. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'MIRseq Mature CNMF subtypes' versus 'CORPUS_INVOLVEMENT'

P value = 0.857 (Fisher's exact test), Q value = 0.93

Table S355. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	95	19
subtype1	28	6
subtype2	31	7
subtype3	36	6

Figure S346. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'MIRseq Mature CNMF subtypes' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.895 (Fisher's exact test), Q value = 0.96

Table S356. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	7	102	2
subtype1	3	37	1
subtype2	2	19	0
subtype3	2	46	1

Figure S347. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'MIRseq Mature CNMF subtypes' versus 'CERVIX_SUV_RESULTS'

P value = 0.18 (Kruskal-Wallis (anova)), Q value = 0.49

Table S357. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

	nPatients	Mean (Std.Dev)
ALL	16	13.0 (7.3)
subtype1	3	16.8 (11.0)
subtype2	7	9.0 (5.4)
subtype3	6	15.7 (6.1)

Figure S348. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

'MIRseq Mature CNMF subtypes' versus 'AGE_AT_DIAGNOSIS'

P value = 0.0326 (Kruskal-Wallis (anova)), Q value = 0.2

Table S358. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	294	48.2 (13.8)
subtype1	104	49.6 (13.1)
subtype2	82	50.1 (14.2)
subtype3	108	45.4 (13.8)

Figure S349. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'MIRseq Mature CNMF subtypes' versus 'CLINICAL_STAGE'

P value = 0.0295 (Fisher's exact test), Q value = 0.2

Table S359. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	5	1	1	37	76	37	5	8	5	7	42	3	41	8	11
subtype1	3	0	0	8	29	16	2	2	1	3	11	2	14	2	7
subtype2	0	1	1	7	27	11	2	3	1	2	6	0	13	4	1
subtype3	2	0	0	22	20	10	1	3	3	2	25	1	14	2	3

Figure S350. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

Clustering Approach #10: 'MIRseq Mature cHierClus subtypes'

Table S360. Description of clustering approach #10: 'MIRseq Mature cHierClus subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	29	81	109	42	33

'MIRseq Mature cHierClus subtypes' versus 'Time to Death'

P value = 0.928 (logrank test), Q value = 0.98

Table S361. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	283	69	0.1 - 210.7 (23.8)
subtype1	28	5	0.5 - 137.2 (19.3)
subtype2	78	20	0.1 - 177.0 (23.8)
subtype3	105	24	0.1 - 210.7 (24.3)
subtype4	40	11	0.4 - 154.3 (23.6)
subtype5	32	9	0.5 - 97.0 (19.6)

Figure S351. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.00654 (Kruskal-Wallis (anova)), Q value = 0.1

Table S362. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	293	48.2 (13.8)
subtype1	29	48.7 (10.1)
subtype2	81	50.6 (14.5)
subtype3	108	49.9 (14.2)
subtype4	42	40.8 (11.7)
subtype5	33	45.9 (13.2)

Figure S352. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.262 (Fisher's exact test), Q value = 0.57

Table S363. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	136	68	19	9
subtype1	13	12	1	1
subtype2	42	13	3	2
subtype3	43	31	10	4
subtype4	18	6	4	1
subtype5	20	6	1	1

Figure S353. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGY_T_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.0778 (Fisher's exact test), Q value = 0.32

Table S364. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	130	57
subtype1	18	2
subtype2	42	17
subtype3	37	26
subtype4	15	7
subtype5	18	5

Figure S354. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_N_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.282 (Fisher's exact test), Q value = 0.6

Table S365. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	112	9
subtype1	15	3
subtype2	40	1
subtype3	35	4
subtype4	14	1
subtype5	8	0

Figure S355. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_M_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'RADIATION_THERAPY'

P value = 0.927 (Fisher's exact test), Q value = 0.98

Table S366. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	53	124
subtype1	7	14
subtype2	11	23
subtype3	19	53
subtype4	9	19
subtype5	7	15

Figure S356. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'RADIATION_THERAPY'

'MIRseq Mature cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 1e-05 (Fisher's exact test), Q value = 0.00043

Table S367. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

nPatients	ADENOSQUAMOUS	CERVICAL SQUAMOUS CELL CARCINOMA	ENDOCERVICAL ADENOCARCINOMA OF THE USUAL TYPE	ENDOCERVICAL TYPE OF ADENOCARCINOMA	ENDOMETRIOID ADENOCARCINOMA OF ENDOCERVIX	MUCINOUS ADENOCARCINOMA OF ENDOCERVICAL TYPE
ALL	6	242	6	20	3	17
subtype1	3	15	1	5	2	3
subtype2	1	78	0	1	0	1
subtype3	1	105	1	2	0	0
subtype4	0	42	0	0	0	0
subtype5	1	2	4	12	1	13

Figure S357. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'HISTOLOGICAL_TYPE'

'MIRseq Mature cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.0646 (Kruskal-Wallis (anova)), Q value = 0.29

Table S368. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	88	17.2 (14.0)
subtype1	5	12.2 (9.9)
subtype2	23	20.7 (10.4)
subtype3	36	19.4 (17.1)
subtype4	13	10.5 (9.9)
subtype5	11	12.9 (12.4)

Figure S358. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'NUMBER_PACK_YEARS_SMOKED'

'MIRseq Mature cHierClus subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.0318 (Kruskal-Wallis (anova)), Q value = 0.2

Table S369. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	155	1.0 (2.3)
subtype1	15	0.1 (0.5)
subtype2	51	0.9 (1.9)
subtype3	52	1.6 (3.1)
subtype4	15	0.3 (0.6)
subtype5	22	0.8 (2.4)

Figure S359. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_OF_LYMPH_NODES'

'MIRseq Mature cHierClus subtypes' versus 'RACE'

P value = 0.0739 (Fisher's exact test), Q value = 0.31

Table S370. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	NATIVE HAWAIIAN OR OTHER PACIFIC ISLANDER	WHITE
ALL	7	19	27	2	205
subtype1	1	4	2	0	17
subtype2	1	9	4	0	59
subtype3	3	4	13	1	74
subtype4	2	2	7	1	27
subtype5	0	0	1	0	28

Figure S360. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'RACE'

'MIRseq Mature cHierClus subtypes' versus 'ETHNICITY'

P value = 0.217 (Fisher's exact test), Q value = 0.55

Table S371. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #11: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	24	161
subtype1	3	16
subtype2	3	55
subtype3	10	52
subtype4	5	22
subtype5	3	16

Figure S361. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #11: 'ETHNICITY'

'MIRseq Mature cHierClus subtypes' versus 'WEIGHT_KG_AT_DIAGNOSIS'

P value = 0.479 (Kruskal-Wallis (anova)), Q value = 0.74

Table S372. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	266	72.9 (21.4)
subtype1	28	73.5 (18.5)
subtype2	72	72.0 (28.3)
subtype3	102	72.9 (19.9)
subtype4	36	74.8 (16.7)
subtype5	28	72.4 (14.8)

Figure S362. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #12: 'WEIGHT_KG_AT_DIAGNOSIS'

'MIRseq Mature cHierClus subtypes' versus 'TUMOR_STATUS'

P value = 0.642 (Fisher's exact test), Q value = 0.84

Table S373. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

nPatients	TUMOR FREE	WITH TUMOR
ALL	194	75
subtype1	17	11
subtype2	54	21
subtype3	73	27
subtype4	29	8
subtype5	21	8

Figure S363. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #13: 'TUMOR_STATUS'

'MIRseq Mature cHierClus subtypes' versus 'NEOPLASM_HISTOLOGIC_GRADE'

P value = 6e-05 (Fisher's exact test), Q value = 0.0021

Table S374. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

nPatients	G1	G2	G3	G4	GX
ALL	16	128	118	1	23
subtype1	1	6	18	0	4
subtype2	7	39	32	0	2
subtype3	3	52	39	1	10
subtype4	0	10	24	0	5
subtype5	5	21	5	0	2

Figure S364. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #14: 'NEOPLASM_HISTOLOGIC_GRADE'

'MIRseq Mature cHierClus subtypes' versus 'TOBACCO_SMOKING_YEAR_STOPPED'

P value = 0.479 (Kruskal-Wallis (anova)), Q value = 0.74

Table S375. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

	nPatients	Mean (Std.Dev)
ALL	41	2000.0 (13.9)
subtype1	2	2003.5 (4.9)
subtype2	9	1996.1 (11.6)
subtype3	18	2000.8 (16.5)
subtype4	6	2002.0 (11.5)
subtype5	6	2000.0 (15.0)

Figure S365. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #15: 'TOBACCO_SMOKING_YEAR_STOPPED'

'MIRseq Mature cHierClus subtypes' versus 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

P value = 0.0646 (Kruskal-Wallis (anova)), Q value = 0.29

Table S376. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	88	17.2 (14.0)
subtype1	5	12.2 (9.9)
subtype2	23	20.7 (10.4)
subtype3	36	19.4 (17.1)
subtype4	13	10.5 (9.9)
subtype5	11	12.9 (12.4)

Figure S366. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #16: 'TOBACCO_SMOKING_PACK_YEARS_SMOKED'

'MIRseq Mature cHierClus subtypes' versus 'TOBACCO_SMOKING_HISTORY'

P value = 0.31 (Kruskal-Wallis (anova)), Q value = 0.63

Table S377. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

	nPatients	Mean (Std.Dev)
ALL	255	1.8 (1.1)
subtype1	25	1.4 (0.9)
subtype2	68	1.8 (1.1)
subtype3	98	1.9 (1.2)
subtype4	33	1.8 (1.2)
subtype5	31	1.8 (1.2)

Figure S367. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #17: 'TOBACCO_SMOKING_HISTORY'

'MIRseq Mature cHierClus subtypes' versus 'AGEBEGANSMOKINGINYEARS'

P value = 0.4 (Kruskal-Wallis (anova)), Q value = 0.7

Table S378. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

	nPatients	Mean (Std.Dev)
ALL	80	20.9 (7.6)
subtype1	4	26.5 (6.6)
subtype2	19	20.2 (4.8)
subtype3	34	21.3 (8.8)
subtype4	12	20.3 (9.8)
subtype5	11	19.5 (4.5)

Figure S368. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #18: 'AGEBEGANSMOKINGINYEARS'

'MIRseq Mature cHierClus subtypes' versus 'RADIATION_THERAPY_STATUS'

P value = 0.0332 (Fisher's exact test), Q value = 0.2

Table S379. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

nPatients	COMPLETED AS PLANNED	TREATMENT NOT COMPLETED
ALL	27	3
subtype1	0	0
subtype2	8	1
subtype3	16	0
subtype4	2	2
subtype5	1	0

Figure S369. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #19: 'RADIATION_THERAPY_STATUS'

'MIRseq Mature cHierClus subtypes' versus 'PREGNANCIES_COUNT_TOTAL'

P value = 0.0979 (Kruskal-Wallis (anova)), Q value = 0.37

Table S380. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

	nPatients	Mean (Std.Dev)
ALL	257	3.6 (2.5)
subtype1	27	2.7 (2.1)
subtype2	70	3.7 (2.1)
subtype3	96	3.9 (2.9)
subtype4	35	3.4 (2.6)
subtype5	29	3.1 (2.1)

Figure S370. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #20: 'PREGNANCIES_COUNT_TOTAL'

'MIRseq Mature cHierClus subtypes' versus 'PREGNANCIES_COUNT_STILLBIRTH'

P value = 0.657 (Kruskal-Wallis (anova)), Q value = 0.84

Table S381. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

	nPatients	Mean (Std.Dev)
ALL	107	0.1 (0.3)
subtype1	7	0.0 (0.0)
subtype2	38	0.1 (0.5)
subtype3	38	0.1 (0.2)
subtype4	10	0.0 (0.0)
subtype5	14	0.0 (0.0)

Figure S371. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #21: 'PREGNANCIES_COUNT_STILLBIRTH'

'MIRseq Mature cHierClus subtypes' versus 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

P value = 0.159 (Kruskal-Wallis (anova)), Q value = 0.48

Table S382. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	143	0.6 (1.0)
subtype1	9	0.4 (0.5)
subtype2	47	0.3 (0.5)
subtype3	54	0.8 (1.3)
subtype4	16	0.6 (0.7)
subtype5	17	0.7 (1.0)

Figure S372. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #22: 'PREGNANCY_SPONTANEOUS_ABORTION_COUNT'

'MIRseq Mature cHierClus subtypes' versus 'PREGNANCIES_COUNT_LIVE_BIRTH'

P value = 0.235 (Kruskal-Wallis (anova)), Q value = 0.55

Table S383. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	252	2.8 (2.0)
subtype1	24	2.4 (1.8)
subtype2	73	2.7 (1.4)
subtype3	93	3.1 (2.2)
subtype4	33	2.9 (2.4)
subtype5	29	2.4 (2.1)

Figure S373. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #23: 'PREGNANCIES_COUNT_LIVE_BIRTH'

'MIRseq Mature cHierClus subtypes' versus 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

P value = 0.323 (Kruskal-Wallis (anova)), Q value = 0.64

Table S384. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

	nPatients	Mean (Std.Dev)
ALL	116	0.8 (1.8)
subtype1	8	1.0 (1.4)
subtype2	41	1.1 (2.2)
subtype3	40	0.6 (1.7)
subtype4	12	0.8 (1.0)
subtype5	15	0.4 (0.6)

Figure S374. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #24: 'PREGNANCY_THERAPEUTIC_ABORTION_COUNT'

'MIRseq Mature cHierClus subtypes' versus 'PREGNANCIES_COUNT_ECTOPIC'

P value = 0.426 (Kruskal-Wallis (anova)), Q value = 0.72

Table S385. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

	nPatients	Mean (Std.Dev)
ALL	112	0.1 (0.3)
subtype1	7	0.0 (0.0)
subtype2	40	0.1 (0.4)
subtype3	41	0.2 (0.4)
subtype4	10	0.0 (0.0)
subtype5	14	0.1 (0.3)

Figure S375. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #25: 'PREGNANCIES_COUNT_ECTOPIC'

'MIRseq Mature cHierClus subtypes' versus 'POS_LYMPH_NODE_LOCATION'

P value = 0.261 (Fisher's exact test), Q value = 0.57

Table S386. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

nPatients	MACROSCOPIC PARAMETRIAL INVOLVEMENT	MICROSCOPIC PARAMETRIAL INVOLVEMENT	OTHER LOCATION, SPECIFY	POSITIVE BLADDER MARGIN	POSITIVE VAGINAL MARGIN
ALL	1	7	39	1	12
subtype1	1	0	1	0	1
subtype2	0	1	15	1	5
subtype3	0	5	12	0	5
subtype4	0	1	5	0	0
subtype5	0	0	6	0	1

Figure S376. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #26: 'POS_LYMPH_NODE_LOCATION'

'MIRseq Mature cHierClus subtypes' versus 'MENOPAUSE_STATUS'

P value = 0.29 (Fisher's exact test), Q value = 0.61

Table S387. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

nPatients	INDETERMINATE (NEITHER PRE OR POSTMENOPAUSAL)	PERI (6-12 MONTHS SINCE LAST MENSTRUAL PERIOD)	POST (PRIOR BILATERAL OVARIECTOMY OR >12 MO SINCE LMP WITH NO PRIOR HYSTERECTOMY)	PRE (<6 MONTHS SINCE LMP AND NO PRIOR BILATERAL OVARIECTOMY AND NOT ON ESTROGEN REPLACEMENT)
ALL	3	25	78	120
subtype1	0	3	6	16
subtype2	1	5	29	27
subtype3	1	11	30	43
subtype4	1	3	4	19
subtype5	0	3	9	15

Figure S377. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #27: 'MENOPAUSE_STATUS'

'MIRseq Mature cHierClus subtypes' versus 'LYMPHOVASCULAR_INVOLVEMENT'

P value = 0.112 (Fisher's exact test), Q value = 0.4

Table S388. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	69	74
subtype1	6	12
subtype2	25	27
subtype3	19	25
subtype4	6	6
subtype5	13	4

Figure S378. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #28: 'LYMPHOVASCULAR_INVOLVEMENT'

'MIRseq Mature cHierClus subtypes' versus 'LYMPH_NODES_EXAMINED_HE_COUNT'

P value = 0.0318 (Kruskal-Wallis (anova)), Q value = 0.2

Table S389. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

	nPatients	Mean (Std.Dev)
ALL	155	1.0 (2.3)
subtype1	15	0.1 (0.5)
subtype2	51	0.9 (1.9)
subtype3	52	1.6 (3.1)
subtype4	15	0.3 (0.6)
subtype5	22	0.8 (2.4)

Figure S379. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #29: 'LYMPH_NODES_EXAMINED_HE_COUNT'

'MIRseq Mature cHierClus subtypes' versus 'LYMPH_NODES_EXAMINED'

P value = 0.316 (Kruskal-Wallis (anova)), Q value = 0.63

Table S390. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

	nPatients	Mean (Std.Dev)
ALL	172	22.7 (12.6)
subtype1	18	19.1 (9.9)
subtype2	57	24.1 (14.2)
subtype3	56	20.9 (12.2)
subtype4	18	24.9 (13.2)
subtype5	23	24.7 (10.7)

Figure S380. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #30: 'LYMPH_NODES_EXAMINED'

'MIRseq Mature cHierClus subtypes' versus 'KERATINIZATION_SQUAMOUS_CELL'

P value = 0.0641 (Fisher's exact test), Q value = 0.29

Table S391. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

nPatients	KERATINIZING SQUAMOUS CELL CARCINOMA	NON-KERATINIZING SQUAMOUS CELL CARCINOMA
ALL	52	114
subtype1	0	10
subtype2	23	35
subtype3	19	46
subtype4	10	19
subtype5	0	4

Figure S381. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #31: 'KERATINIZATION_SQUAMOUS_CELL'

'MIRseq Mature cHierClus subtypes' versus 'INITIAL_PATHOLOGIC_DX_YEAR'

P value = 0.0245 (Kruskal-Wallis (anova)), Q value = 0.2

Table S392. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

	nPatients	Mean (Std.Dev)
ALL	292	2008.3 (4.8)
subtype1	29	2010.7 (2.5)
subtype2	80	2007.4 (5.3)
subtype3	109	2008.7 (4.6)
subtype4	42	2007.3 (5.2)
subtype5	32	2008.6 (4.2)

Figure S382. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #32: 'INITIAL_PATHOLOGIC_DX_YEAR'

'MIRseq Mature cHierClus subtypes' versus 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

P value = 0.662 (Fisher's exact test), Q value = 0.84

Table S393. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

nPatients	CURRENT USER	FORMER USER	NEVER USED
ALL	15	54	85
subtype1	4	6	8
subtype2	2	13	22
subtype3	4	21	35
subtype4	2	6	11
subtype5	3	8	9

Figure S383. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #33: 'HISTORY_HORMONAL_CONTRACEPTIVES_USE'

'MIRseq Mature cHierClus subtypes' versus 'HEIGHT_CM_AT_DIAGNOSIS'

P value = 0.437 (Kruskal-Wallis (anova)), Q value = 0.72

Table S394. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	253	160.8 (7.3)
subtype1	28	160.1 (4.9)
subtype2	66	160.1 (7.8)
subtype3	98	160.6 (7.2)
subtype4	34	162.0 (7.7)
subtype5	27	162.4 (8.2)

Figure S384. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #34: 'HEIGHT_CM_AT_DIAGNOSIS'

'MIRseq Mature cHierClus subtypes' versus 'CORPUS_INVOLVEMENT'

P value = 0.359 (Fisher's exact test), Q value = 0.66

Table S395. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

nPatients	ABSENT	PRESENT
ALL	95	19
subtype1	10	4
subtype2	34	9
subtype3	28	5
subtype4	8	0
subtype5	15	1

Figure S385. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #35: 'CORPUS_INVOLVEMENT'

'MIRseq Mature cHierClus subtypes' versus 'CHEMO_CONCURRENT_TYPE'

P value = 0.08 (Fisher's exact test), Q value = 0.32

Table S396. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

nPatients	CARBOPLATIN	CISPLATIN	OTHER
ALL	7	102	2
subtype1	2	9	0
subtype2	2	19	0
subtype3	1	48	0
subtype4	1	16	1
subtype5	1	10	1

Figure S386. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #36: 'CHEMO_CONCURRENT_TYPE'

'MIRseq Mature cHierClus subtypes' versus 'CERVIX_SUV_RESULTS'

P value = 0.19 (Kruskal-Wallis (anova)), Q value = 0.51

Table S397. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

	nPatients	Mean (Std.Dev)
ALL	16	13.0 (7.3)
subtype1	1	7.1 (NA)
subtype2	7	9.5 (5.2)
subtype3	5	15.8 (8.6)
subtype4	3	18.2 (7.3)

Figure S387. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #37: 'CERVIX_SUV_RESULTS'

'MIRseq Mature cHierClus subtypes' versus 'AGE_AT_DIAGNOSIS'

P value = 0.00599 (Kruskal-Wallis (anova)), Q value = 0.1

Table S398. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

	nPatients	Mean (Std.Dev)
ALL	294	48.2 (13.8)
subtype1	29	48.7 (10.1)
subtype2	81	50.6 (14.5)
subtype3	109	49.9 (14.2)
subtype4	42	40.8 (11.7)
subtype5	33	45.9 (13.2)

Figure S388. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #38: 'AGE_AT_DIAGNOSIS'

'MIRseq Mature cHierClus subtypes' versus 'CLINICAL_STAGE'

P value = 0.514 (Fisher's exact test), Q value = 0.78

Table S399. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IA2	STAGE IB	STAGE IB1	STAGE IB2	STAGE II	STAGE IIA	STAGE IIA1	STAGE IIA2	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IVA	STAGE IVB
ALL	5	1	1	37	76	37	5	8	5	7	42	3	41	8	11
subtype1	1	0	0	2	8	5	0	2	1	2	4	0	2	0	2
subtype2	0	1	1	14	22	9	3	2	2	2	7	0	13	3	1
subtype3	3	0	0	11	24	10	2	1	2	3	20	1	18	3	6
subtype4	1	0	0	7	7	7	0	2	0	0	8	2	5	1	1
subtype5	0	0	0	3	15	6	0	1	0	0	3	0	3	1	1

Figure S389. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #39: 'CLINICAL_STAGE'

Methods & Data

Input

Cluster data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/GDAC_mergedClustering/CESC-TP/22541064/CESC-TP.mergedcluster.txt
Clinical data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/Append_Data/CESC-TP/22489370/CESC-TP.merged_data.txt
Number of patients = 307
Number of clustering approaches = 10
Number of selected clinical features = 39
Exclude small clusters that include fewer than K patients, K = 3

Clustering approaches

CNMF clustering

consensus non-negative matrix factorization clustering approach (Brunet et al. 2004)

Consensus hierarchical clustering

Resampling-based clustering method (Monti et al. 2003)

Survival analysis

For survival clinical features, the Kaplan-Meier survival curves of tumors with and without gene mutations were plotted and the statistical significance P values were estimated by logrank test (Bland and Altman 2004) using the 'survdiff' function in R

Fisher's exact test

For binary clinical features, two-tailed Fisher's exact tests (Fisher 1922) were used to estimate the P values using the 'fisher.test' function in R

Q value calculation

For multiple hypothesis correction, Q value is the False Discovery Rate (FDR) analogue of the P value (Benjamini and Hochberg 1995), defined as the minimum FDR at which the test may be called significant. We used the 'Benjamini and Hochberg' method of 'p.adjust' function in R to convert P values into Q values.

References

[1] Brunet et al., Metagenes and molecular pattern discovery using matrix factorization, PNAS 101(12):4164-9 (2004)

[2] Monti et al., Consensus Clustering: A Resampling-Based Method for Class Discovery and Visualization of Gene Expression Microarray Data, Machine Learning 52(1):91-118 (2003)

[3] Bland and Altman, Statistics notes: The logrank test, BMJ 328(7447):1073 (2004)

[4] Fisher, R.A., On the interpretation of chi-square from contingency tables, and the calculation of P, Journal of the Royal Statistical Society 85(1):87-94 (1922)

[5] Benjamini and Hochberg, Controlling the false discovery rate: a practical and powerful approach to multiple testing, Journal of the Royal Statistical Society Series B 59:289-300 (1995)

Made with Nozzle