Correlation between aggregated molecular cancer subtypes and selected clinical features

Lung Adenocarcinoma (Primary solid tumor)

21 August 2015 | analyses__2015_08_21

Maintainer Information

Citation Information

Maintained by TCGA GDAC Team (Broad Institute/MD Anderson Cancer Center/Harvard Medical School)

Cite as Broad Institute TCGA Genome Data Analysis Center (2015): Correlation between aggregated molecular cancer subtypes and selected clinical features. Broad Institute of MIT and Harvard. doi:10.7908/C15T3JQ2

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 12 different clustering approaches and 15 clinical features across 519 patients, 58 significant findings detected with P value < 0.05 and Q value < 0.25.

CNMF clustering analysis on array-based mRNA expression data identified 4 subtypes that correlate to 'PATHOLOGIC_STAGE'.
Consensus hierarchical clustering analysis on array-based mRNA expression data identified 3 subtypes that correlate to 'PATHOLOGIC_STAGE' and 'RACE'.
4 subtypes identified in current cancer cohort by 'Copy Number Ratio CNMF subtypes'. These subtypes correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'GENDER', and 'NUMBER_PACK_YEARS_SMOKED'.
3 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. These subtypes correlate to 'PATHOLOGIC_STAGE', 'PATHOLOGY_T_STAGE', 'GENDER', 'RADIATION_THERAPY', 'NUMBER_PACK_YEARS_SMOKED', and 'ETHNICITY'.
CNMF clustering analysis on RPPA data identified 5 subtypes that correlate to 'YEARS_TO_BIRTH', 'PATHOLOGY_M_STAGE', 'HISTOLOGICAL_TYPE', 'YEAR_OF_TOBACCO_SMOKING_ONSET', and 'RESIDUAL_TUMOR'.
Consensus hierarchical clustering analysis on RPPA data identified 3 subtypes that correlate to 'HISTOLOGICAL_TYPE', 'YEAR_OF_TOBACCO_SMOKING_ONSET', and 'ETHNICITY'.
CNMF clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'PATHOLOGIC_STAGE', 'PATHOLOGY_T_STAGE', 'PATHOLOGY_N_STAGE', 'GENDER', 'RADIATION_THERAPY', and 'HISTOLOGICAL_TYPE'.
Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 7 subtypes that correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'PATHOLOGIC_STAGE', 'PATHOLOGY_T_STAGE', 'PATHOLOGY_N_STAGE', 'GENDER', 'RADIATION_THERAPY', 'HISTOLOGICAL_TYPE', 'NUMBER_PACK_YEARS_SMOKED', 'YEAR_OF_TOBACCO_SMOKING_ONSET', and 'ETHNICITY'.
3 subtypes identified in current cancer cohort by 'MIRSEQ CNMF'. These subtypes correlate to 'YEARS_TO_BIRTH', 'PATHOLOGIC_STAGE', 'PATHOLOGY_T_STAGE', 'HISTOLOGICAL_TYPE', and 'YEAR_OF_TOBACCO_SMOKING_ONSET'.
4 subtypes identified in current cancer cohort by 'MIRSEQ CHIERARCHICAL'. These subtypes correlate to 'YEARS_TO_BIRTH', 'GENDER', 'HISTOLOGICAL_TYPE', 'NUMBER_PACK_YEARS_SMOKED', and 'YEAR_OF_TOBACCO_SMOKING_ONSET'.
3 subtypes identified in current cancer cohort by 'MIRseq Mature CNMF subtypes'. These subtypes correlate to 'PATHOLOGY_T_STAGE', 'HISTOLOGICAL_TYPE', and 'NUMBER_PACK_YEARS_SMOKED'.
3 subtypes identified in current cancer cohort by 'MIRseq Mature cHierClus subtypes'. These subtypes correlate to 'YEARS_TO_BIRTH', 'PATHOLOGY_T_STAGE', 'GENDER', 'HISTOLOGICAL_TYPE', and 'YEAR_OF_TOBACCO_SMOKING_ONSET'.

Results

Overview of the results

Table 1. Get Full Table Overview of the association between subtypes identified by 12 different clustering approaches and 15 clinical features. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, 58 significant findings detected.


Clinical Features	Statistical Tests	mRNA CNMF subtypes	mRNA cHierClus subtypes	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	0.944 (1.00)	0.912 (0.983)	0.0133 (0.0601)	0.131 (0.314)	0.997 (1.00)	0.29 (0.522)	0.00175 (0.0131)	2.96e-08 (5.32e-06)	0.0533 (0.163)	0.818 (0.92)	0.177 (0.394)	0.0556 (0.167)
YEARS TO BIRTH	Kruskal-Wallis (anova)	0.65 (0.8)	0.513 (0.684)	0.00156 (0.0128)	0.206 (0.418)	0.00173 (0.0131)	0.092 (0.254)	0.0479 (0.15)	0.000674 (0.00933)	0.0102 (0.0469)	0.00155 (0.0128)	0.657 (0.8)	2.92e-06 (0.000263)
PATHOLOGIC STAGE	Fisher's exact test	0.0454 (0.149)	0.0363 (0.128)	0.12 (0.292)	0.00037 (0.00605)	0.333 (0.576)	0.663 (0.8)	0.00325 (0.0196)	0.0028 (0.0187)	0.00182 (0.0131)	0.659 (0.8)	0.432 (0.668)	0.492 (0.684)
PATHOLOGY T STAGE	Fisher's exact test	0.512 (0.684)	0.276 (0.504)	0.115 (0.286)	0.00102 (0.0115)	0.205 (0.418)	0.105 (0.267)	0.00156 (0.0128)	0.00114 (0.0121)	0.0176 (0.0736)	0.133 (0.316)	0.0152 (0.0667)	0.00927 (0.0451)
PATHOLOGY N STAGE	Fisher's exact test	0.482 (0.684)	0.452 (0.668)	0.249 (0.477)	0.0623 (0.184)	0.434 (0.668)	0.483 (0.684)	0.0266 (0.103)	0.00307 (0.0196)	0.276 (0.504)	0.716 (0.837)	0.188 (0.399)	0.0721 (0.209)
PATHOLOGY M STAGE	Fisher's exact test	0.502 (0.684)	0.492 (0.684)	0.187 (0.399)	0.26 (0.493)	0.0484 (0.15)	0.791 (0.901)	0.445 (0.668)	0.507 (0.684)	0.546 (0.712)	0.856 (0.939)	0.435 (0.668)	0.577 (0.737)
GENDER	Fisher's exact test	0.273 (0.504)	0.669 (0.803)	0.0189 (0.0775)	2e-05 (6e-04)	0.157 (0.362)	0.539 (0.708)	1e-05 (0.00045)	2e-05 (6e-04)	0.476 (0.684)	0.00477 (0.0268)	0.959 (1.00)	0.00029 (0.00522)
RADIATION THERAPY	Fisher's exact test	1 (1.00)	1 (1.00)	0.592 (0.75)	0.0322 (0.118)	0.202 (0.418)	0.105 (0.267)	0.0268 (0.103)	0.0234 (0.0935)	0.241 (0.472)	0.69 (0.817)	0.829 (0.921)	0.187 (0.399)
KARNOFSKY PERFORMANCE SCORE	Kruskal-Wallis (anova)			0.359 (0.604)	0.402 (0.64)	0.306 (0.546)	0.824 (0.921)	0.193 (0.404)	0.384 (0.623)	0.277 (0.504)	0.651 (0.8)	0.118 (0.292)	0.835 (0.922)
HISTOLOGICAL TYPE	Fisher's exact test	0.431 (0.668)	0.396 (0.636)	0.382 (0.623)	0.14 (0.326)	0.00724 (0.0372)	0.00138 (0.0128)	1e-05 (0.00045)	0.00013 (0.0026)	0.00096 (0.0115)	0.00123 (0.0123)	4e-05 (9e-04)	3e-05 (0.000771)
NUMBER PACK YEARS SMOKED	Kruskal-Wallis (anova)	0.46 (0.668)	0.362 (0.604)	0.0303 (0.114)	0.00402 (0.0233)	0.78 (0.895)	0.982 (1.00)	0.0931 (0.254)	0.0416 (0.141)	0.0797 (0.224)	0.0407 (0.141)	0.00889 (0.0444)	0.224 (0.442)
YEAR OF TOBACCO SMOKING ONSET	Kruskal-Wallis (anova)	0.715 (0.837)	0.676 (0.806)	0.173 (0.388)	0.888 (0.963)	0.000782 (0.0101)	0.00327 (0.0196)	0.355 (0.604)	0.000498 (0.00747)	0.017 (0.073)	0.00525 (0.0286)	0.506 (0.684)	0.00232 (0.0161)
RESIDUAL TUMOR	Fisher's exact test	0.535 (0.708)	0.458 (0.668)	0.878 (0.957)	0.8 (0.906)	0.00716 (0.0372)	0.776 (0.895)	0.645 (0.8)	0.244 (0.472)	0.45 (0.668)	0.953 (1.00)	0.161 (0.366)	0.0973 (0.261)
RACE	Fisher's exact test	0.0784 (0.224)	0.00957 (0.0453)	0.561 (0.721)	0.183 (0.399)	0.332 (0.576)	0.363 (0.604)	0.972 (1.00)	0.601 (0.757)	0.44 (0.668)	0.317 (0.56)	0.949 (1.00)	0.102 (0.267)
ETHNICITY	Fisher's exact test	1 (1.00)	1 (1.00)	0.221 (0.442)	0.0463 (0.149)	0.55 (0.713)	0.0458 (0.149)	0.496 (0.684)	0.0361 (0.128)	0.369 (0.609)	0.758 (0.88)	0.101 (0.267)	0.455 (0.668)

Clustering Approach #1: 'mRNA CNMF subtypes'

Table S1. Description of clustering approach #1: 'mRNA CNMF subtypes'

Cluster Labels	1	2	3	4
Number of samples	5	9	12	6

'mRNA CNMF subtypes' versus 'Time to Death'

P value = 0.944 (logrank test), Q value = 1

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

	nPatients	nDeath	Duration Range (Median), Month
ALL	31	7	0.5 - 56.8 (34.1)
subtype1	4	1	6.0 - 48.6 (38.2)
subtype2	9	2	4.0 - 56.8 (38.2)
subtype3	12	2	0.5 - 44.9 (22.7)
subtype4	6	2	20.1 - 45.2 (30.9)

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

'mRNA CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.65 (Kruskal-Wallis (anova)), Q value = 0.8

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

	nPatients	Mean (Std.Dev)
ALL	30	65.7 (10.8)
subtype1	4	58.5 (15.5)
subtype2	9	65.0 (9.1)
subtype3	12	67.1 (11.1)
subtype4	5	69.4 (9.0)

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

'mRNA CNMF subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.0454 (Fisher's exact test), Q value = 0.15

Table S4. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE IA	STAGE IB	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IV
ALL	12	11	1	3	3	2
subtype1	3	0	0	1	0	1
subtype2	4	4	0	0	1	0
subtype3	3	7	0	1	0	1
subtype4	2	0	1	1	2	0

Figure S3. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'mRNA CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.512 (Fisher's exact test), Q value = 0.68

Table S5. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3
ALL	12	19	1
subtype1	3	2	0
subtype2	4	4	1
subtype3	4	8	0
subtype4	1	5	0

Figure S4. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'mRNA CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.482 (Fisher's exact test), Q value = 0.68

Table S6. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	23	4	4
subtype1	3	1	1
subtype2	8	1	0
subtype3	9	1	1
subtype4	3	1	2

Figure S5. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'mRNA CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.502 (Fisher's exact test), Q value = 0.68

Table S7. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	30	2
subtype1	4	1
subtype2	9	0
subtype3	11	1
subtype4	6	0

Figure S6. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'mRNA CNMF subtypes' versus 'GENDER'

P value = 0.273 (Fisher's exact test), Q value = 0.5

Table S8. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	18	14
subtype1	3	2
subtype2	3	6
subtype3	9	3
subtype4	3	3

Figure S7. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'mRNA CNMF subtypes' versus 'RADIATION_THERAPY'

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

Table S9. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	18	1
subtype1	3	0
subtype2	6	0
subtype3	6	1
subtype4	3	0

Figure S8. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'mRNA CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.431 (Fisher's exact test), Q value = 0.67

Table S10. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG CLEAR CELL ADENOCARCINOMA
ALL	1	30	1
subtype1	0	4	1
subtype2	0	9	0
subtype3	1	11	0
subtype4	0	6	0

Figure S9. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'mRNA CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.46 (Kruskal-Wallis (anova)), Q value = 0.67

Table S11. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	20	41.1 (15.0)
subtype1	2	29.0 (12.7)
subtype2	9	47.0 (15.5)
subtype3	6	37.0 (13.1)
subtype4	3	40.0 (17.3)

Figure S10. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'mRNA CNMF subtypes' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.715 (Kruskal-Wallis (anova)), Q value = 0.84

Table S12. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	19	1968.5 (11.4)
subtype1	2	1977.0 (18.4)
subtype2	6	1971.2 (14.2)
subtype3	6	1965.8 (8.2)
subtype4	5	1965.2 (9.8)

Figure S11. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'mRNA CNMF subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.535 (Fisher's exact test), Q value = 0.71

Table S13. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R2	RX
ALL	26	1	2
subtype1	3	0	1
subtype2	7	1	0
subtype3	10	0	1
subtype4	6	0	0

Figure S12. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'mRNA CNMF subtypes' versus 'RACE'

P value = 0.0784 (Fisher's exact test), Q value = 0.22

Table S14. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #14: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	1	26
subtype1	1	0	3
subtype2	0	0	7
subtype3	0	0	12
subtype4	1	1	4

Figure S13. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #14: 'RACE'

'mRNA CNMF subtypes' versus 'ETHNICITY'

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

Table S15. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	1	28
subtype1	0	4
subtype2	0	7
subtype3	1	11
subtype4	0	6

Figure S14. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #2: 'mRNA cHierClus subtypes'

Table S16. Description of clustering approach #2: 'mRNA cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	14	11	7

'mRNA cHierClus subtypes' versus 'Time to Death'

P value = 0.912 (logrank test), Q value = 0.98

Table S17. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	31	7	0.5 - 56.8 (34.1)
subtype1	13	2	0.5 - 47.0 (26.6)
subtype2	11	3	4.0 - 56.8 (34.1)
subtype3	7	2	20.1 - 48.6 (38.7)

Figure S15. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'mRNA cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.513 (Kruskal-Wallis (anova)), Q value = 0.68

Table S18. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	30	65.7 (10.8)
subtype1	14	66.7 (10.5)
subtype2	11	62.7 (12.0)
subtype3	5	69.4 (9.0)

Figure S16. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'mRNA cHierClus subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.0363 (Fisher's exact test), Q value = 0.13

Table S19. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE IA	STAGE IB	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IV
ALL	12	11	1	3	3	2
subtype1	3	8	0	1	0	2
subtype2	6	3	0	1	1	0
subtype3	3	0	1	1	2	0

Figure S17. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'mRNA cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.276 (Fisher's exact test), Q value = 0.5

Table S20. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3
ALL	12	19	1
subtype1	4	10	0
subtype2	6	4	1
subtype3	2	5	0

Figure S18. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'mRNA cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.452 (Fisher's exact test), Q value = 0.67

Table S21. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	23	4	4
subtype1	10	1	2
subtype2	9	2	0
subtype3	4	1	2

Figure S19. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'mRNA cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.492 (Fisher's exact test), Q value = 0.68

Table S22. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	30	2
subtype1	12	2
subtype2	11	0
subtype3	7	0

Figure S20. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'mRNA cHierClus subtypes' versus 'GENDER'

P value = 0.669 (Fisher's exact test), Q value = 0.8

Table S23. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	18	14
subtype1	9	5
subtype2	5	6
subtype3	4	3

Figure S21. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'mRNA cHierClus subtypes' versus 'RADIATION_THERAPY'

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

Table S24. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	18	1
subtype1	7	1
subtype2	7	0
subtype3	4	0

Figure S22. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'mRNA cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.396 (Fisher's exact test), Q value = 0.64

Table S25. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG CLEAR CELL ADENOCARCINOMA
ALL	1	30	1
subtype1	1	13	0
subtype2	0	11	0
subtype3	0	6	1

Figure S23. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'mRNA cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.362 (Kruskal-Wallis (anova)), Q value = 0.6

Table S26. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	20	41.1 (15.0)
subtype1	8	35.4 (12.2)
subtype2	9	46.7 (16.1)
subtype3	3	40.0 (17.3)

Figure S24. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'mRNA cHierClus subtypes' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.676 (Kruskal-Wallis (anova)), Q value = 0.81

Table S27. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	19	1968.5 (11.4)
subtype1	7	1969.9 (13.0)
subtype2	6	1970.5 (12.9)
subtype3	6	1965.0 (8.8)

Figure S25. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'mRNA cHierClus subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.458 (Fisher's exact test), Q value = 0.67

Table S28. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R2	RX
ALL	26	1	2
subtype1	10	0	2
subtype2	10	1	0
subtype3	6	0	0

Figure S26. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'mRNA cHierClus subtypes' versus 'RACE'

P value = 0.00957 (Fisher's exact test), Q value = 0.045

Table S29. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #14: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	1	26
subtype1	0	0	13
subtype2	0	0	9
subtype3	2	1	4

Figure S27. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #14: 'RACE'

'mRNA cHierClus subtypes' versus 'ETHNICITY'

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

Table S30. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	1	28
subtype1	1	12
subtype2	0	9
subtype3	0	7

Figure S28. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #3: 'Copy Number Ratio CNMF subtypes'

Table S31. Description of clustering approach #3: 'Copy Number Ratio CNMF subtypes'

Cluster Labels	1	2	3	4
Number of samples	69	228	78	140

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

P value = 0.0133 (logrank test), Q value = 0.06

Table S32. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	492	167	0.0 - 238.3 (20.0)
subtype1	65	20	0.1 - 129.5 (18.6)
subtype2	218	60	0.1 - 238.3 (20.6)
subtype3	75	32	0.0 - 86.1 (18.7)
subtype4	134	55	0.1 - 232.2 (21.4)

Figure S29. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'Copy Number Ratio CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.00156 (Kruskal-Wallis (anova)), Q value = 0.013

Table S33. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	484	65.2 (10.0)
subtype1	65	64.4 (11.0)
subtype2	213	67.2 (8.9)
subtype3	74	64.9 (10.8)
subtype4	132	62.7 (10.2)

Figure S30. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.12 (Fisher's exact test), Q value = 0.29

Table S34. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	5	135	139	1	52	72	72	11	26
subtype1	0	22	16	0	4	7	13	2	5
subtype2	2	74	63	1	20	28	27	3	8
subtype3	1	15	18	0	8	16	11	2	7
subtype4	2	24	42	0	20	21	21	4	6

Figure S31. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.115 (Fisher's exact test), Q value = 0.29

Table S35. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	170	278	46	19
subtype1	24	36	5	4
subtype2	90	112	16	9
subtype3	19	45	11	2
subtype4	37	85	14	4

Figure S32. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.249 (Fisher's exact test), Q value = 0.48

Table S36. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	332	96	76
subtype1	48	8	12
subtype2	154	40	27
subtype3	49	16	12
subtype4	81	32	25

Figure S33. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.187 (Fisher's exact test), Q value = 0.4

Table S37. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	349	25
subtype1	52	5
subtype2	156	7
subtype3	51	7
subtype4	90	6

Figure S34. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'GENDER'

P value = 0.0189 (Fisher's exact test), Q value = 0.078

Table S38. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	275	240
subtype1	28	41
subtype2	128	100
subtype3	35	43
subtype4	84	56

Figure S35. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'Copy Number Ratio CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.592 (Fisher's exact test), Q value = 0.75

Table S39. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	360	54
subtype1	52	9
subtype2	163	20
subtype3	49	10
subtype4	96	15

Figure S36. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'Copy Number Ratio CNMF subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.359 (Kruskal-Wallis (anova)), Q value = 0.6

Table S40. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	132	78.8 (28.6)
subtype1	17	74.7 (36.6)
subtype2	50	85.0 (22.2)
subtype3	24	76.2 (28.6)
subtype4	41	74.4 (31.4)

Figure S37. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'Copy Number Ratio CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.382 (Fisher's exact test), Q value = 0.62

Table S41. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	19	106	320	5	19	2	3	2	23	1	5	10
subtype1	6	14	41	0	1	0	0	0	7	0	0	0
subtype2	7	50	132	3	11	1	1	2	10	1	3	7
subtype3	2	14	54	0	1	1	0	0	2	0	1	3
subtype4	4	28	93	2	6	0	2	0	4	0	1	0

Figure S38. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'Copy Number Ratio CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.0303 (Kruskal-Wallis (anova)), Q value = 0.11

Table S42. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	355	41.6 (27.3)
subtype1	54	40.7 (28.1)
subtype2	152	38.2 (26.3)
subtype3	55	41.9 (25.5)
subtype4	94	47.5 (28.8)

Figure S39. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'Copy Number Ratio CNMF subtypes' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.173 (Kruskal-Wallis (anova)), Q value = 0.39

Table S43. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	276	1965.0 (12.6)
subtype1	43	1965.9 (13.6)
subtype2	119	1963.5 (12.1)
subtype3	41	1968.3 (12.6)
subtype4	73	1965.1 (12.5)

Figure S40. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'Copy Number Ratio CNMF subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.878 (Fisher's exact test), Q value = 0.96

Table S44. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	343	12	4	26
subtype1	43	2	0	5
subtype2	146	3	2	12
subtype3	55	2	1	3
subtype4	99	5	1	6

Figure S41. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'Copy Number Ratio CNMF subtypes' versus 'RACE'

P value = 0.561 (Fisher's exact test), Q value = 0.72

Table S45. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #14: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	1	7	51	389
subtype1	0	2	5	53
subtype2	0	2	20	180
subtype3	0	1	10	55
subtype4	1	2	16	101

Figure S42. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #14: 'RACE'

'Copy Number Ratio CNMF subtypes' versus 'ETHNICITY'

P value = 0.221 (Fisher's exact test), Q value = 0.44

Table S46. Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	7	382
subtype1	1	51
subtype2	2	177
subtype3	3	55
subtype4	1	99

Figure S43. Get High-res Image Clustering Approach #3: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #4: 'METHLYATION CNMF'

Table S47. Description of clustering approach #4: 'METHLYATION CNMF'

Cluster Labels	1	2	3
Number of samples	145	148	164

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.131 (logrank test), Q value = 0.31

Table S48. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	435	145	0.1 - 238.3 (19.5)
subtype1	137	35	0.1 - 238.3 (20.6)
subtype2	141	58	0.1 - 232.2 (19.0)
subtype3	157	52	0.1 - 163.1 (18.1)

Figure S44. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'YEARS_TO_BIRTH'

P value = 0.206 (Kruskal-Wallis (anova)), Q value = 0.42

Table S49. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	428	65.0 (10.2)
subtype1	134	66.0 (9.7)
subtype2	140	63.9 (10.4)
subtype3	154	65.2 (10.3)

Figure S45. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'METHLYATION CNMF' versus 'PATHOLOGIC_STAGE'

P value = 0.00037 (Fisher's exact test), Q value = 0.0061

Table S50. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	5	120	124	1	49	62	64	9	21
subtype1	1	57	34	1	16	17	12	1	5
subtype2	4	35	37	0	18	14	26	4	10
subtype3	0	28	53	0	15	31	26	4	6

Figure S46. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_T_STAGE'

P value = 0.00102 (Fisher's exact test), Q value = 0.011

Table S51. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	154	243	41	16
subtype1	68	62	10	3
subtype2	48	83	11	5
subtype3	38	98	20	8

Figure S47. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_N_STAGE'

P value = 0.0623 (Fisher's exact test), Q value = 0.18

Table S52. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	298	83	66
subtype1	101	25	12
subtype2	92	24	30
subtype3	105	34	24

Figure S48. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_M_STAGE'

P value = 0.26 (Fisher's exact test), Q value = 0.49

Table S53. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	295	19
subtype1	84	4
subtype2	98	10
subtype3	113	5

Figure S49. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'METHLYATION CNMF' versus 'GENDER'

P value = 2e-05 (Fisher's exact test), Q value = 6e-04

Table S54. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	243	214
subtype1	97	48
subtype2	59	89
subtype3	87	77

Figure S50. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #7: 'GENDER'

'METHLYATION CNMF' versus 'RADIATION_THERAPY'

P value = 0.0322 (Fisher's exact test), Q value = 0.12

Table S55. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	328	49
subtype1	106	16
subtype2	100	23
subtype3	122	10

Figure S51. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #8: 'RADIATION_THERAPY'

'METHLYATION CNMF' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.402 (Kruskal-Wallis (anova)), Q value = 0.64

Table S56. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	118	79.8 (28.3)
subtype1	32	84.7 (25.3)
subtype2	46	76.1 (31.3)
subtype3	40	80.2 (27.1)

Figure S52. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'METHLYATION CNMF' versus 'HISTOLOGICAL_TYPE'

P value = 0.14 (Fisher's exact test), Q value = 0.33

Table S57. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	19	91	280	5	19	1	2	2	22	1	5	10
subtype1	5	23	88	4	8	0	1	2	7	1	0	6
subtype2	8	29	92	0	3	1	0	0	9	0	3	3
subtype3	6	39	100	1	8	0	1	0	6	0	2	1

Figure S53. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'METHLYATION CNMF' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.00402 (Kruskal-Wallis (anova)), Q value = 0.023

Table S58. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	313	40.7 (27.3)
subtype1	91	37.4 (24.6)
subtype2	110	47.4 (29.6)
subtype3	112	36.8 (25.9)

Figure S54. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'METHLYATION CNMF' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.888 (Kruskal-Wallis (anova)), Q value = 0.96

Table S59. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	251	1965.3 (12.5)
subtype1	72	1964.7 (11.4)
subtype2	88	1965.8 (12.9)
subtype3	91	1965.2 (13.0)

Figure S55. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'METHLYATION CNMF' versus 'RESIDUAL_TUMOR'

P value = 0.8 (Fisher's exact test), Q value = 0.91

Table S60. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	293	10	1	23
subtype1	87	2	1	9
subtype2	94	4	0	6
subtype3	112	4	0	8

Figure S56. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'METHLYATION CNMF' versus 'RACE'

P value = 0.183 (Fisher's exact test), Q value = 0.4

Table S61. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #14: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	6	49	352
subtype1	0	18	114
subtype2	2	19	114
subtype3	4	12	124

Figure S57. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #14: 'RACE'

'METHLYATION CNMF' versus 'ETHNICITY'

P value = 0.0463 (Fisher's exact test), Q value = 0.15

Table S62. Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	7	341
subtype1	2	118
subtype2	5	108
subtype3	0	115

Figure S58. Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #5: 'RPPA CNMF subtypes'

Table S63. Description of clustering approach #5: 'RPPA CNMF subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	91	23	67	75	109

'RPPA CNMF subtypes' versus 'Time to Death'

P value = 0.997 (logrank test), Q value = 1

Table S64. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	347	133	0.1 - 238.3 (21.5)
subtype1	85	31	0.1 - 123.6 (22.8)
subtype2	22	7	0.1 - 73.9 (18.8)
subtype3	65	32	0.3 - 163.1 (23.1)
subtype4	70	27	0.8 - 97.7 (21.7)
subtype5	105	36	0.1 - 238.3 (20.5)

Figure S59. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.00173 (Kruskal-Wallis (anova)), Q value = 0.013

Table S65. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	344	64.4 (9.8)
subtype1	83	63.7 (10.3)
subtype2	23	65.3 (9.6)
subtype3	66	64.9 (9.3)
subtype4	68	60.6 (9.4)
subtype5	104	66.9 (9.4)

Figure S60. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RPPA CNMF subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.333 (Fisher's exact test), Q value = 0.58

Table S66. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	2	96	95	1	38	50	57	8	18
subtype1	1	31	23	0	10	13	11	1	1
subtype2	0	8	6	0	1	1	4	0	3
subtype3	0	11	20	0	9	9	9	3	6
subtype4	0	15	20	1	5	14	13	3	4
subtype5	1	31	26	0	13	13	20	1	4

Figure S61. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.205 (Fisher's exact test), Q value = 0.42

Table S67. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	117	203	28	16
subtype1	35	50	5	1
subtype2	8	12	1	1
subtype3	15	39	6	7
subtype4	19	46	6	4
subtype5	40	56	10	3

Figure S62. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.434 (Fisher's exact test), Q value = 0.67

Table S68. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	225	72	59
subtype1	54	20	13
subtype2	14	4	3
subtype3	44	14	8
subtype4	41	20	14
subtype5	72	14	21

Figure S63. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.0484 (Fisher's exact test), Q value = 0.15

Table S69. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	259	17
subtype1	67	1
subtype2	15	3
subtype3	51	6
subtype4	50	4
subtype5	76	3

Figure S64. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RPPA CNMF subtypes' versus 'GENDER'

P value = 0.157 (Fisher's exact test), Q value = 0.36

Table S70. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	196	169
subtype1	57	34
subtype2	15	8
subtype3	32	35
subtype4	35	40
subtype5	57	52

Figure S65. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'RPPA CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.202 (Fisher's exact test), Q value = 0.42

Table S71. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	263	41
subtype1	65	11
subtype2	15	3
subtype3	54	4
subtype4	49	13
subtype5	80	10

Figure S66. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'RPPA CNMF subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.306 (Kruskal-Wallis (anova)), Q value = 0.55

Table S72. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	97	77.0 (30.5)
subtype1	24	69.2 (38.7)
subtype2	3	90.0 (10.0)
subtype3	23	87.0 (18.4)
subtype4	24	78.8 (30.0)
subtype5	23	71.7 (31.3)

Figure S67. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'RPPA CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.00724 (Fisher's exact test), Q value = 0.037

Table S73. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	13	73	233	5	13	1	2	2	13	1	4	5
subtype1	2	12	67	1	3	0	0	1	4	0	0	1
subtype2	2	8	10	0	1	0	0	0	2	0	0	0
subtype3	2	21	35	0	5	0	1	0	2	0	0	1
subtype4	3	17	50	0	0	1	0	0	0	0	4	0
subtype5	4	15	71	4	4	0	1	1	5	1	0	3

Figure S68. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'RPPA CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.78 (Kruskal-Wallis (anova)), Q value = 0.89

Table S74. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	259	41.6 (27.9)
subtype1	47	41.8 (24.1)
subtype2	19	41.3 (32.5)
subtype3	54	45.5 (31.0)
subtype4	61	41.9 (29.1)
subtype5	78	38.7 (25.9)

Figure S69. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'RPPA CNMF subtypes' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.000782 (Kruskal-Wallis (anova)), Q value = 0.01

Table S75. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	205	1964.9 (13.0)
subtype1	33	1965.3 (13.2)
subtype2	14	1961.7 (10.8)
subtype3	45	1959.4 (11.5)
subtype4	48	1971.1 (13.5)
subtype5	65	1964.6 (12.3)

Figure S70. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'RPPA CNMF subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.00716 (Fisher's exact test), Q value = 0.037

Table S76. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	235	9	3	21
subtype1	59	2	0	5
subtype2	12	0	3	5
subtype3	50	3	0	4
subtype4	50	1	0	3
subtype5	64	3	0	4

Figure S71. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'RPPA CNMF subtypes' versus 'RACE'

P value = 0.332 (Fisher's exact test), Q value = 0.58

Table S77. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #14: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	5	39	274
subtype1	1	6	73
subtype2	1	3	13
subtype3	1	6	52
subtype4	1	12	50
subtype5	1	12	86

Figure S72. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #14: 'RACE'

'RPPA CNMF subtypes' versus 'ETHNICITY'

P value = 0.55 (Fisher's exact test), Q value = 0.71

Table S78. Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	6	260
subtype1	1	68
subtype2	0	15
subtype3	0	42
subtype4	3	51
subtype5	2	84

Figure S73. Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #6: 'RPPA cHierClus subtypes'

Table S79. Description of clustering approach #6: 'RPPA cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	167	135	63

'RPPA cHierClus subtypes' versus 'Time to Death'

P value = 0.29 (logrank test), Q value = 0.52

Table S80. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	347	133	0.1 - 238.3 (21.5)
subtype1	159	64	0.1 - 164.1 (22.7)
subtype2	127	51	0.4 - 97.7 (20.1)
subtype3	61	18	0.1 - 238.3 (21.6)

Figure S74. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.092 (Kruskal-Wallis (anova)), Q value = 0.25

Table S81. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	344	64.4 (9.8)
subtype1	160	65.2 (9.3)
subtype2	124	62.7 (10.2)
subtype3	60	65.8 (10.2)

Figure S75. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RPPA cHierClus subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.663 (Fisher's exact test), Q value = 0.8

Table S82. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	2	96	95	1	38	50	57	8	18
subtype1	1	51	40	0	16	19	27	4	9
subtype2	1	28	38	1	13	26	20	3	5
subtype3	0	17	17	0	9	5	10	1	4

Figure S76. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.105 (Fisher's exact test), Q value = 0.27

Table S83. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	117	203	28	16
subtype1	61	86	9	10
subtype2	34	83	15	3
subtype3	22	34	4	3

Figure S77. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.483 (Fisher's exact test), Q value = 0.68

Table S84. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	225	72	59
subtype1	102	31	27
subtype2	81	33	21
subtype3	42	8	11

Figure S78. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.791 (Fisher's exact test), Q value = 0.9

Table S85. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	259	17
subtype1	121	9
subtype2	97	5
subtype3	41	3

Figure S79. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RPPA cHierClus subtypes' versus 'GENDER'

P value = 0.539 (Fisher's exact test), Q value = 0.71

Table S86. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	196	169
subtype1	91	76
subtype2	68	67
subtype3	37	26

Figure S80. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'RPPA cHierClus subtypes' versus 'RADIATION_THERAPY'

P value = 0.105 (Fisher's exact test), Q value = 0.27

Table S87. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	263	41
subtype1	125	15
subtype2	89	21
subtype3	49	5

Figure S81. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'RPPA cHierClus subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.824 (Kruskal-Wallis (anova)), Q value = 0.92

Table S88. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	97	77.0 (30.5)
subtype1	42	77.4 (30.4)
subtype2	39	77.7 (31.3)
subtype3	16	74.4 (30.5)

Figure S82. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'RPPA cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.00138 (Fisher's exact test), Q value = 0.013

Table S89. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	13	73	233	5	13	1	2	2	13	1	4	5
subtype1	6	41	94	2	8	1	2	1	8	1	0	3
subtype2	3	21	103	0	1	0	0	1	1	0	4	1
subtype3	4	11	36	3	4	0	0	0	4	0	0	1

Figure S83. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'RPPA cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.982 (Kruskal-Wallis (anova)), Q value = 1

Table S90. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	259	41.6 (27.9)
subtype1	112	42.4 (30.0)
subtype2	95	42.1 (28.1)
subtype3	52	39.0 (22.4)

Figure S84. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'RPPA cHierClus subtypes' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.00327 (Kruskal-Wallis (anova)), Q value = 0.02

Table S91. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	205	1964.9 (13.0)
subtype1	86	1961.1 (12.2)
subtype2	74	1968.4 (13.5)
subtype3	45	1966.3 (12.2)

Figure S85. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'RPPA cHierClus subtypes' versus 'RESIDUAL_TUMOR'

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

Table S92. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	235	9	3	21
subtype1	110	6	2	13
subtype2	88	2	1	5
subtype3	37	1	0	3

Figure S86. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'RPPA cHierClus subtypes' versus 'RACE'

P value = 0.363 (Fisher's exact test), Q value = 0.6

Table S93. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #14: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	5	39	274
subtype1	2	14	126
subtype2	3	19	94
subtype3	0	6	54

Figure S87. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #14: 'RACE'

'RPPA cHierClus subtypes' versus 'ETHNICITY'

P value = 0.0458 (Fisher's exact test), Q value = 0.15

Table S94. Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	6	260
subtype1	0	117
subtype2	4	95
subtype3	2	48

Figure S88. Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #7: 'RNAseq CNMF subtypes'

Table S95. Description of clustering approach #7: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	167	197	150

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.00175 (logrank test), Q value = 0.013

Table S96. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	491	167	0.0 - 238.3 (19.9)
subtype1	158	37	0.1 - 238.3 (20.4)
subtype2	189	81	0.0 - 221.3 (19.1)
subtype3	144	49	0.1 - 232.2 (19.7)

Figure S89. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.0479 (Kruskal-Wallis (anova)), Q value = 0.15

Table S97. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	483	65.3 (10.0)
subtype1	157	66.9 (9.3)
subtype2	184	64.2 (10.6)
subtype3	142	64.9 (9.6)

Figure S90. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RNAseq CNMF subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.00325 (Fisher's exact test), Q value = 0.02

Table S98. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	5	133	139	1	52	71	73	11	27
subtype1	2	63	46	0	14	15	19	1	5
subtype2	3	33	54	1	21	35	29	7	14
subtype3	0	37	39	0	17	21	25	3	8

Figure S91. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.00156 (Fisher's exact test), Q value = 0.013

Table S99. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	168	277	47	19
subtype1	77	72	13	4
subtype2	49	117	19	10
subtype3	42	88	15	5

Figure S92. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.0266 (Fisher's exact test), Q value = 0.1

Table S100. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	330	96	76
subtype1	119	25	17
subtype2	113	47	33
subtype3	98	24	26

Figure S93. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.445 (Fisher's exact test), Q value = 0.67

Table S101. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	346	25
subtype1	109	5
subtype2	129	12
subtype3	108	8

Figure S94. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 1e-05 (Fisher's exact test), Q value = 0.00045

Table S102. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	276	238
subtype1	112	55
subtype2	113	84
subtype3	51	99

Figure S95. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'RNAseq CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.0268 (Fisher's exact test), Q value = 0.1

Table S103. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	356	55
subtype1	127	10
subtype2	129	27
subtype3	100	18

Figure S96. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'RNAseq CNMF subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.193 (Kruskal-Wallis (anova)), Q value = 0.4

Table S104. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	130	78.3 (28.9)
subtype1	38	83.7 (26.3)
subtype2	58	74.7 (30.8)
subtype3	34	78.5 (28.2)

Figure S97. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'RNAseq CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 1e-05 (Fisher's exact test), Q value = 0.00045

Table S105. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	19	106	319	5	19	2	3	2	23	1	5	10
subtype1	7	35	90	5	14	0	1	1	8	1	1	4
subtype2	3	33	147	0	3	2	2	0	4	0	3	0
subtype3	9	38	82	0	2	0	0	1	11	0	1	6

Figure S98. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'RNAseq CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.0931 (Kruskal-Wallis (anova)), Q value = 0.25

Table S106. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	351	41.6 (27.3)
subtype1	97	37.8 (25.7)
subtype2	132	44.9 (27.7)
subtype3	122	41.0 (27.9)

Figure S99. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'RNAseq CNMF subtypes' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.355 (Kruskal-Wallis (anova)), Q value = 0.6

Table S107. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	273	1965.0 (12.5)
subtype1	80	1964.1 (10.3)
subtype2	98	1966.3 (13.9)
subtype3	95	1964.3 (12.7)

Figure S100. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'RNAseq CNMF subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.645 (Fisher's exact test), Q value = 0.8

Table S108. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	342	12	4	25
subtype1	112	3	1	11
subtype2	127	6	3	8
subtype3	103	3	0	6

Figure S101. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'RNAseq CNMF subtypes' versus 'RACE'

P value = 0.972 (Fisher's exact test), Q value = 1

Table S109. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #14: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	1	8	51	387
subtype1	0	3	18	131
subtype2	1	2	19	148
subtype3	0	3	14	108

Figure S102. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #14: 'RACE'

'RNAseq CNMF subtypes' versus 'ETHNICITY'

P value = 0.496 (Fisher's exact test), Q value = 0.68

Table S110. Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	7	381
subtype1	1	136
subtype2	4	147
subtype3	2	98

Figure S103. Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #8: 'RNAseq cHierClus subtypes'

Table S111. Description of clustering approach #8: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3	4	5	6	7
Number of samples	79	112	97	48	52	49	77

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 2.96e-08 (logrank test), Q value = 5.3e-06

Table S112. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	491	167	0.0 - 238.3 (19.9)
subtype1	75	22	0.1 - 163.1 (19.8)
subtype2	107	30	0.4 - 97.7 (19.1)
subtype3	93	38	0.1 - 238.3 (20.0)
subtype4	45	16	0.4 - 129.5 (26.5)
subtype5	49	13	0.1 - 88.0 (16.0)
subtype6	48	31	0.0 - 55.9 (15.1)
subtype7	74	17	0.1 - 164.1 (22.7)

Figure S104. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.000674 (Kruskal-Wallis (anova)), Q value = 0.0093

Table S113. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	483	65.3 (10.0)
subtype1	75	68.3 (9.2)
subtype2	103	63.5 (10.8)
subtype3	90	65.5 (9.7)
subtype4	46	67.7 (7.7)
subtype5	49	60.7 (9.6)
subtype6	47	66.0 (9.9)
subtype7	73	65.5 (10.2)

Figure S105. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RNAseq cHierClus subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.0028 (Fisher's exact test), Q value = 0.019

Table S114. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	5	133	139	1	52	71	73	11	27
subtype1	0	30	22	0	8	6	9	1	3
subtype2	1	16	35	0	9	24	18	3	6
subtype3	3	26	27	1	14	7	11	2	6
subtype4	0	16	16	0	4	5	3	2	2
subtype5	0	10	15	0	7	5	12	1	2
subtype6	0	7	7	0	3	12	12	2	6
subtype7	1	28	17	0	7	12	8	0	2

Figure S106. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.00114 (Fisher's exact test), Q value = 0.012

Table S115. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	168	277	47	19
subtype1	36	36	4	3
subtype2	24	69	13	5
subtype3	39	51	3	3
subtype4	17	24	4	3
subtype5	11	35	5	1
subtype6	8	29	8	3
subtype7	33	33	10	1

Figure S107. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.00307 (Fisher's exact test), Q value = 0.02

Table S116. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	330	96	76
subtype1	57	12	9
subtype2	66	25	19
subtype3	60	23	12
subtype4	37	8	2
subtype5	36	3	12
subtype6	22	12	14
subtype7	52	13	8

Figure S108. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.507 (Fisher's exact test), Q value = 0.68

Table S117. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	346	25
subtype1	52	3
subtype2	79	4
subtype3	55	6
subtype4	38	2
subtype5	38	2
subtype6	37	6
subtype7	47	2

Figure S109. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 2e-05 (Fisher's exact test), Q value = 6e-04

Table S118. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	276	238
subtype1	63	16
subtype2	59	53
subtype3	52	45
subtype4	16	32
subtype5	23	29
subtype6	19	30
subtype7	44	33

Figure S110. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'RNAseq cHierClus subtypes' versus 'RADIATION_THERAPY'

P value = 0.0234 (Fisher's exact test), Q value = 0.094

Table S119. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	356	55
subtype1	62	2
subtype2	76	10
subtype3	64	15
subtype4	37	3
subtype5	36	5
subtype6	31	9
subtype7	50	11

Figure S111. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'RNAseq cHierClus subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.384 (Kruskal-Wallis (anova)), Q value = 0.62

Table S120. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	130	78.3 (28.9)
subtype1	15	87.3 (25.2)
subtype2	28	77.5 (28.5)
subtype3	32	79.1 (29.1)
subtype4	11	89.1 (13.0)
subtype5	14	66.4 (36.7)
subtype6	10	72.0 (33.3)
subtype7	20	77.0 (29.6)

Figure S112. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'RNAseq cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.00013 (Fisher's exact test), Q value = 0.0026

Table S121. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	19	106	319	5	19	2	3	2	23	1	5	10
subtype1	5	16	45	1	6	0	1	0	5	0	0	0
subtype2	2	18	81	0	3	1	2	0	4	0	1	0
subtype3	3	15	71	1	2	1	0	0	2	0	2	0
subtype4	2	19	18	0	1	0	0	0	7	0	0	1
subtype5	4	10	36	0	1	0	0	0	0	0	0	1
subtype6	1	12	31	0	0	0	0	0	2	0	1	2
subtype7	2	16	37	3	6	0	0	2	3	1	1	6

Figure S113. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'RNAseq cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.0416 (Kruskal-Wallis (anova)), Q value = 0.14

Table S122. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	351	41.6 (27.3)
subtype1	46	35.4 (25.7)
subtype2	78	41.1 (25.7)
subtype3	64	49.9 (31.5)
subtype4	37	40.4 (30.5)
subtype5	42	37.3 (26.3)
subtype6	39	45.9 (24.1)
subtype7	45	38.2 (24.2)

Figure S114. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'RNAseq cHierClus subtypes' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.000498 (Kruskal-Wallis (anova)), Q value = 0.0075

Table S123. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	273	1965.0 (12.5)
subtype1	37	1963.0 (10.3)
subtype2	59	1967.5 (14.2)
subtype3	50	1963.5 (12.3)
subtype4	29	1956.9 (11.4)
subtype5	34	1970.3 (9.7)
subtype6	29	1965.1 (13.3)
subtype7	35	1966.2 (11.2)

Figure S115. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'RNAseq cHierClus subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.244 (Fisher's exact test), Q value = 0.47

Table S124. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	342	12	4	25
subtype1	54	4	0	2
subtype2	77	1	2	3
subtype3	58	3	2	9
subtype4	28	0	0	3
subtype5	37	1	0	2
subtype6	35	3	0	2
subtype7	53	0	0	4

Figure S116. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'RNAseq cHierClus subtypes' versus 'RACE'

P value = 0.601 (Fisher's exact test), Q value = 0.76

Table S125. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #14: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	1	8	51	387
subtype1	0	2	6	62
subtype2	1	3	9	84
subtype3	0	1	14	70
subtype4	0	0	5	37
subtype5	0	0	5	38
subtype6	0	2	2	38
subtype7	0	0	10	58

Figure S117. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #14: 'RACE'

'RNAseq cHierClus subtypes' versus 'ETHNICITY'

P value = 0.0361 (Fisher's exact test), Q value = 0.13

Table S126. Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	7	381
subtype1	1	60
subtype2	0	85
subtype3	4	71
subtype4	0	34
subtype5	0	36
subtype6	2	31
subtype7	0	64

Figure S118. Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #9: 'MIRSEQ CNMF'

Table S127. Description of clustering approach #9: 'MIRSEQ CNMF'

Cluster Labels	1	2	3
Number of samples	190	227	95

'MIRSEQ CNMF' versus 'Time to Death'

P value = 0.0533 (logrank test), Q value = 0.16

Table S128. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	489	166	0.0 - 238.3 (20.0)
subtype1	181	53	0.1 - 164.1 (19.8)
subtype2	219	74	0.0 - 232.2 (20.2)
subtype3	89	39	0.1 - 238.3 (19.7)

Figure S119. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CNMF' versus 'YEARS_TO_BIRTH'

P value = 0.0102 (Kruskal-Wallis (anova)), Q value = 0.047

Table S129. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	482	65.3 (10.0)
subtype1	180	67.0 (9.4)
subtype2	217	64.0 (10.3)
subtype3	85	65.1 (9.7)

Figure S120. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRSEQ CNMF' versus 'PATHOLOGIC_STAGE'

P value = 0.00182 (Fisher's exact test), Q value = 0.013

Table S130. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	5	134	139	1	52	70	73	11	25
subtype1	1	71	40	0	19	25	22	1	9
subtype2	4	47	69	0	22	34	30	8	13
subtype3	0	16	30	1	11	11	21	2	3

Figure S121. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_T_STAGE'

P value = 0.0176 (Fisher's exact test), Q value = 0.074

Table S131. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	169	274	47	19
subtype1	83	85	16	6
subtype2	60	133	23	9
subtype3	26	56	8	4

Figure S122. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_N_STAGE'

P value = 0.276 (Fisher's exact test), Q value = 0.5

Table S132. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	329	95	76
subtype1	124	39	21
subtype2	149	38	36
subtype3	56	18	19

Figure S123. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_M_STAGE'

P value = 0.546 (Fisher's exact test), Q value = 0.71

Table S133. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	346	23
subtype1	130	8
subtype2	157	13
subtype3	59	2

Figure S124. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRSEQ CNMF' versus 'GENDER'

P value = 0.476 (Fisher's exact test), Q value = 0.68

Table S134. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	273	239
subtype1	108	82
subtype2	116	111
subtype3	49	46

Figure S125. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #7: 'GENDER'

'MIRSEQ CNMF' versus 'RADIATION_THERAPY'

P value = 0.241 (Fisher's exact test), Q value = 0.47

Table S135. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	355	54
subtype1	140	16
subtype2	147	29
subtype3	68	9

Figure S126. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #8: 'RADIATION_THERAPY'

'MIRSEQ CNMF' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.277 (Kruskal-Wallis (anova)), Q value = 0.5

Table S136. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	128	78.5 (29.1)
subtype1	43	81.4 (27.8)
subtype2	66	79.8 (28.0)
subtype3	19	67.4 (34.1)

Figure S127. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'MIRSEQ CNMF' versus 'HISTOLOGICAL_TYPE'

P value = 0.00096 (Fisher's exact test), Q value = 0.011

Table S137. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	19	105	319	5	19	2	2	2	23	1	5	10
subtype1	10	45	99	4	12	1	1	2	8	1	2	5
subtype2	8	52	148	0	4	1	0	0	9	0	3	2
subtype3	1	8	72	1	3	0	1	0	6	0	0	3

Figure S128. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'MIRSEQ CNMF' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.0797 (Kruskal-Wallis (anova)), Q value = 0.22

Table S138. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	351	41.8 (27.4)
subtype1	125	38.2 (25.8)
subtype2	168	43.2 (27.9)
subtype3	58	45.3 (28.8)

Figure S129. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'MIRSEQ CNMF' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.017 (Kruskal-Wallis (anova)), Q value = 0.073

Table S139. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	273	1964.8 (12.5)
subtype1	104	1962.3 (11.9)
subtype2	129	1967.1 (12.7)
subtype3	40	1963.5 (12.1)

Figure S130. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'MIRSEQ CNMF' versus 'RESIDUAL_TUMOR'

P value = 0.45 (Fisher's exact test), Q value = 0.67

Table S140. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	339	12	4	26
subtype1	122	2	2	10
subtype2	162	7	2	9
subtype3	55	3	0	7

Figure S131. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'MIRSEQ CNMF' versus 'RACE'

P value = 0.44 (Fisher's exact test), Q value = 0.67

Table S141. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #14: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	8	51	386
subtype1	1	18	152
subtype2	5	25	158
subtype3	2	8	76

Figure S132. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #14: 'RACE'

'MIRSEQ CNMF' versus 'ETHNICITY'

P value = 0.369 (Fisher's exact test), Q value = 0.61

Table S142. Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	7	379
subtype1	1	144
subtype2	5	161
subtype3	1	74

Figure S133. Get High-res Image Clustering Approach #9: 'MIRSEQ CNMF' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #10: 'MIRSEQ CHIERARCHICAL'

Table S143. Description of clustering approach #10: 'MIRSEQ CHIERARCHICAL'

Cluster Labels	1	2	3	4
Number of samples	72	198	142	100

'MIRSEQ CHIERARCHICAL' versus 'Time to Death'

P value = 0.818 (logrank test), Q value = 0.92

Table S144. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	489	166	0.0 - 238.3 (20.0)
subtype1	69	22	0.1 - 129.5 (21.4)
subtype2	191	66	0.0 - 238.3 (19.0)
subtype3	134	44	0.1 - 164.1 (19.1)
subtype4	95	34	0.1 - 232.2 (21.9)

Figure S134. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CHIERARCHICAL' versus 'YEARS_TO_BIRTH'

P value = 0.00155 (Kruskal-Wallis (anova)), Q value = 0.013

Table S145. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	482	65.3 (10.0)
subtype1	68	68.1 (7.8)
subtype2	188	63.7 (10.0)
subtype3	133	66.9 (9.8)
subtype4	93	64.1 (10.8)

Figure S135. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGIC_STAGE'

P value = 0.659 (Fisher's exact test), Q value = 0.8

Table S146. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	5	134	139	1	52	70	73	11	25
subtype1	0	28	17	0	3	10	10	1	3
subtype2	3	40	60	0	20	30	30	6	9
subtype3	1	42	35	1	16	19	16	2	8
subtype4	1	24	27	0	13	11	17	2	5

Figure S136. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_T_STAGE'

P value = 0.133 (Fisher's exact test), Q value = 0.32

Table S147. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	169	274	47	19
subtype1	31	33	4	4
subtype2	50	120	19	8
subtype3	55	67	14	4
subtype4	33	54	10	3

Figure S137. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_N_STAGE'

P value = 0.716 (Fisher's exact test), Q value = 0.84

Table S148. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	329	95	76
subtype1	48	14	7
subtype2	128	36	32
subtype3	89	29	18
subtype4	64	16	19

Figure S138. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_M_STAGE'

P value = 0.856 (Fisher's exact test), Q value = 0.94

Table S149. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	346	23
subtype1	50	2
subtype2	145	9
subtype3	89	7
subtype4	62	5

Figure S139. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'GENDER'

P value = 0.00477 (Fisher's exact test), Q value = 0.027

Table S150. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	273	239
subtype1	35	37
subtype2	113	85
subtype3	86	56
subtype4	39	61

Figure S140. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'GENDER'

'MIRSEQ CHIERARCHICAL' versus 'RADIATION_THERAPY'

P value = 0.69 (Fisher's exact test), Q value = 0.82

Table S151. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	355	54
subtype1	54	9
subtype2	132	23
subtype3	99	11
subtype4	70	11

Figure S141. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'RADIATION_THERAPY'

'MIRSEQ CHIERARCHICAL' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.651 (Kruskal-Wallis (anova)), Q value = 0.8

Table S152. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	128	78.5 (29.1)
subtype1	14	80.0 (25.7)
subtype2	54	78.7 (31.0)
subtype3	30	81.0 (28.4)
subtype4	30	75.0 (28.5)

Figure S142. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'MIRSEQ CHIERARCHICAL' versus 'HISTOLOGICAL_TYPE'

P value = 0.00123 (Fisher's exact test), Q value = 0.012

Table S153. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	19	105	319	5	19	2	2	2	23	1	5	10
subtype1	6	18	39	0	4	1	0	0	4	0	0	0
subtype2	4	39	136	2	4	0	2	0	10	0	1	0
subtype3	6	30	73	3	10	0	0	2	7	1	2	8
subtype4	3	18	71	0	1	1	0	0	2	0	2	2

Figure S143. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'MIRSEQ CHIERARCHICAL' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.0407 (Kruskal-Wallis (anova)), Q value = 0.14

Table S154. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	351	41.8 (27.4)
subtype1	49	43.3 (29.9)
subtype2	136	41.4 (25.3)
subtype3	92	36.2 (24.3)
subtype4	74	48.4 (31.5)

Figure S144. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'MIRSEQ CHIERARCHICAL' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.00525 (Kruskal-Wallis (anova)), Q value = 0.029

Table S155. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	273	1964.8 (12.5)
subtype1	47	1959.6 (12.7)
subtype2	102	1966.3 (12.0)
subtype3	71	1964.1 (11.7)
subtype4	53	1967.2 (13.1)

Figure S145. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'MIRSEQ CHIERARCHICAL' versus 'RESIDUAL_TUMOR'

P value = 0.953 (Fisher's exact test), Q value = 1

Table S156. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	339	12	4	26
subtype1	41	1	0	4
subtype2	140	6	1	8
subtype3	90	3	2	9
subtype4	68	2	1	5

Figure S146. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'MIRSEQ CHIERARCHICAL' versus 'RACE'

P value = 0.317 (Fisher's exact test), Q value = 0.56

Table S157. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #14: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	8	51	386
subtype1	1	6	58
subtype2	4	18	150
subtype3	2	11	114
subtype4	1	16	64

Figure S147. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #14: 'RACE'

'MIRSEQ CHIERARCHICAL' versus 'ETHNICITY'

P value = 0.758 (Fisher's exact test), Q value = 0.88

Table S158. Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	7	379
subtype1	0	52
subtype2	3	149
subtype3	2	108
subtype4	2	70

Figure S148. Get High-res Image Clustering Approach #10: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #11: 'MIRseq Mature CNMF subtypes'

Table S159. Description of clustering approach #11: 'MIRseq Mature CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	146	197	93

'MIRseq Mature CNMF subtypes' versus 'Time to Death'

P value = 0.177 (logrank test), Q value = 0.39

Table S160. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	414	139	0.1 - 238.3 (19.5)
subtype1	137	53	0.1 - 163.1 (19.3)
subtype2	190	55	0.1 - 232.2 (19.6)
subtype3	87	31	0.1 - 238.3 (20.7)

Figure S149. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.657 (Kruskal-Wallis (anova)), Q value = 0.8

Table S161. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	408	65.1 (10.0)
subtype1	137	65.4 (10.7)
subtype2	188	64.7 (9.7)
subtype3	83	65.4 (9.9)

Figure S150. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.432 (Fisher's exact test), Q value = 0.67

Table S162. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	5	113	114	1	49	60	64	8	21
subtype1	1	35	34	0	19	18	24	4	10
subtype2	4	59	51	1	23	26	22	3	8
subtype3	0	19	29	0	7	16	18	1	3

Figure S151. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.0152 (Fisher's exact test), Q value = 0.067

Table S163. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	147	231	40	15
subtype1	48	73	14	11
subtype2	76	100	18	2
subtype3	23	58	8	2

Figure S152. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.188 (Fisher's exact test), Q value = 0.4

Table S164. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	280	81	65
subtype1	86	34	21
subtype2	138	31	26
subtype3	56	16	18

Figure S153. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.435 (Fisher's exact test), Q value = 0.67

Table S165. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	280	19
subtype1	97	9
subtype2	124	8
subtype3	59	2

Figure S154. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRseq Mature CNMF subtypes' versus 'GENDER'

P value = 0.959 (Fisher's exact test), Q value = 1

Table S166. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	233	203
subtype1	77	69
subtype2	105	92
subtype3	51	42

Figure S155. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'MIRseq Mature CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.829 (Fisher's exact test), Q value = 0.92

Table S167. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	311	49
subtype1	102	15
subtype2	145	22
subtype3	64	12

Figure S156. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'MIRseq Mature CNMF subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.118 (Kruskal-Wallis (anova)), Q value = 0.29

Table S168. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	113	79.6 (28.9)
subtype1	27	67.4 (38.9)
subtype2	66	84.4 (23.3)
subtype3	20	80.0 (26.2)

Figure S157. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'MIRseq Mature CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 4e-05 (Fisher's exact test), Q value = 9e-04

Table S169. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	17	85	267	5	19	1	2	2	22	1	5	10
subtype1	4	38	75	3	11	1	2	1	7	0	0	4
subtype2	13	39	117	0	7	0	0	0	11	1	5	4
subtype3	0	8	75	2	1	0	0	1	4	0	0	2

Figure S158. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'MIRseq Mature CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.00889 (Kruskal-Wallis (anova)), Q value = 0.044

Table S170. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	296	41.3 (27.5)
subtype1	104	36.2 (26.9)
subtype2	139	43.8 (28.0)
subtype3	53	44.9 (26.3)

Figure S159. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'MIRseq Mature CNMF subtypes' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.506 (Kruskal-Wallis (anova)), Q value = 0.68

Table S171. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	235	1965.0 (12.3)
subtype1	77	1964.6 (13.2)
subtype2	121	1966.0 (11.8)
subtype3	37	1962.8 (11.9)

Figure S160. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'MIRseq Mature CNMF subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.161 (Fisher's exact test), Q value = 0.37

Table S172. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	279	9	1	22
subtype1	83	5	1	4
subtype2	139	2	0	15
subtype3	57	2	0	3

Figure S161. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'MIRseq Mature CNMF subtypes' versus 'RACE'

P value = 0.949 (Fisher's exact test), Q value = 1

Table S173. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #14: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	5	49	332
subtype1	1	15	110
subtype2	3	24	148
subtype3	1	10	74

Figure S162. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #14: 'RACE'

'MIRseq Mature CNMF subtypes' versus 'ETHNICITY'

P value = 0.101 (Fisher's exact test), Q value = 0.27

Table S174. Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	6	324
subtype1	4	91
subtype2	2	159
subtype3	0	74

Figure S163. Get High-res Image Clustering Approach #11: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #15: 'ETHNICITY'

Clustering Approach #12: 'MIRseq Mature cHierClus subtypes'

Table S175. Description of clustering approach #12: 'MIRseq Mature cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	103	247	86

'MIRseq Mature cHierClus subtypes' versus 'Time to Death'

P value = 0.0556 (logrank test), Q value = 0.17

Table S176. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	414	139	0.1 - 238.3 (19.5)
subtype1	99	23	0.1 - 164.1 (20.5)
subtype2	233	84	0.1 - 232.2 (18.9)
subtype3	82	32	0.1 - 238.3 (19.0)

Figure S164. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 2.92e-06 (Kruskal-Wallis (anova)), Q value = 0.00026

Table S177. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	408	65.1 (10.0)
subtype1	98	68.3 (8.1)
subtype2	229	62.9 (10.4)
subtype3	81	67.2 (9.7)

Figure S165. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.492 (Fisher's exact test), Q value = 0.68

Table S178. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IV
ALL	5	113	114	1	49	60	64	8	21
subtype1	1	34	23	1	10	16	9	2	6
subtype2	4	55	67	0	31	36	39	4	11
subtype3	0	24	24	0	8	8	16	2	4

Figure S166. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.00927 (Fisher's exact test), Q value = 0.045

Table S179. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	147	231	40	15
subtype1	43	42	13	4
subtype2	74	143	24	5
subtype3	30	46	3	6

Figure S167. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.0721 (Fisher's exact test), Q value = 0.21

Table S180. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2+N3
ALL	280	81	65
subtype1	75	15	8
subtype2	155	46	43
subtype3	50	20	14

Figure S168. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.577 (Fisher's exact test), Q value = 0.74

Table S181. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	280	19
subtype1	66	6
subtype2	161	11
subtype3	53	2

Figure S169. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'GENDER'

P value = 0.00029 (Fisher's exact test), Q value = 0.0052

Table S182. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	233	203
subtype1	52	51
subtype2	119	128
subtype3	62	24

Figure S170. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'MIRseq Mature cHierClus subtypes' versus 'RADIATION_THERAPY'

P value = 0.187 (Fisher's exact test), Q value = 0.4

Table S183. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	311	49
subtype1	82	7
subtype2	172	32
subtype3	57	10

Figure S171. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'MIRseq Mature cHierClus subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.835 (Kruskal-Wallis (anova)), Q value = 0.92

Table S184. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	113	79.6 (28.9)
subtype1	22	87.3 (16.7)
subtype2	75	78.0 (30.4)
subtype3	16	76.2 (34.0)

Figure S172. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #9: 'KARNOFSKY_PERFORMANCE_SCORE'

'MIRseq Mature cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 3e-05 (Fisher's exact test), Q value = 0.00077

Table S185. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

nPatients	LUNG ACINAR ADENOCARCINOMA	LUNG ADENOCARCINOMA MIXED SUBTYPE	LUNG ADENOCARCINOMA- NOT OTHERWISE SPECIFIED (NOS)	LUNG BRONCHIOLOALVEOLAR CARCINOMA MUCINOUS	LUNG BRONCHIOLOALVEOLAR CARCINOMA NONMUCINOUS	LUNG CLEAR CELL ADENOCARCINOMA	LUNG MICROPAPILLARY ADENOCARCINOMA	LUNG MUCINOUS ADENOCARCINOMA	LUNG PAPILLARY ADENOCARCINOMA	LUNG SIGNET RING ADENOCARCINOMA	LUNG SOLID PATTERN PREDOMINANT ADENOCARCINOMA	MUCINOUS (COLLOID) CARCINOMA
ALL	17	85	267	5	19	1	2	2	22	1	5	10
subtype1	5	27	43	3	6	0	0	2	8	1	1	7
subtype2	9	43	169	0	8	1	0	0	10	0	4	3
subtype3	3	15	55	2	5	0	2	0	4	0	0	0

Figure S173. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'HISTOLOGICAL_TYPE'

'MIRseq Mature cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.224 (Kruskal-Wallis (anova)), Q value = 0.44

Table S186. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	296	41.3 (27.5)
subtype1	67	39.6 (29.0)
subtype2	176	43.1 (27.8)
subtype3	53	37.6 (24.5)

Figure S174. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #11: 'NUMBER_PACK_YEARS_SMOKED'

'MIRseq Mature cHierClus subtypes' versus 'YEAR_OF_TOBACCO_SMOKING_ONSET'

P value = 0.00232 (Kruskal-Wallis (anova)), Q value = 0.016

Table S187. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

	nPatients	Mean (Std.Dev)
ALL	235	1965.0 (12.3)
subtype1	53	1961.2 (10.6)
subtype2	138	1967.3 (12.1)
subtype3	44	1962.6 (13.5)

Figure S175. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #12: 'YEAR_OF_TOBACCO_SMOKING_ONSET'

'MIRseq Mature cHierClus subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.0973 (Fisher's exact test), Q value = 0.26

Table S188. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	279	9	1	22
subtype1	65	0	0	9
subtype2	164	5	1	10
subtype3	50	4	0	3

Figure S176. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #13: 'RESIDUAL_TUMOR'

'MIRseq Mature cHierClus subtypes' versus 'RACE'

P value = 0.102 (Fisher's exact test), Q value = 0.27

Table S189. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #14: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	5	49	332
subtype1	2	9	81
subtype2	1	34	181
subtype3	2	6	70

Figure S177. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #14: 'RACE'

'MIRseq Mature cHierClus subtypes' versus 'ETHNICITY'

P value = 0.455 (Fisher's exact test), Q value = 0.67

Table S190. Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #15: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	6	324
subtype1	0	82
subtype2	5	182
subtype3	1	60

Figure S178. Get High-res Image Clustering Approach #12: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #15: 'ETHNICITY'

Methods & Data

Input

Cluster data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/GDAC_mergedClustering/LUAD-TP/20140677/LUAD-TP.mergedcluster.txt
Clinical data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/Append_Data/LUAD-TP/19775350/LUAD-TP.merged_data.txt
Number of patients = 519
Number of clustering approaches = 12
Number of selected clinical features = 15
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.

Download Results

In addition to the links below, the full results of the analysis summarized in this report can also be downloaded programmatically using firehose_get, or interactively from either the Broad GDAC website or TCGA Data Coordination Center Portal.

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