Correlation between aggregated molecular cancer subtypes and selected clinical features

Overview

Introduction

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

Summary

Testing the association between subtypes identified by 10 different clustering approaches and 12 clinical features across 393 patients, 52 significant findings detected with P value < 0.05 and Q value < 0.25.

4 subtypes identified in current cancer cohort by 'Copy Number Ratio CNMF subtypes'. These subtypes correlate to 'PATHOLOGY_N_STAGE' and 'ETHNICITY'.
4 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. These subtypes correlate to 'YEARS_TO_BIRTH', 'NEOPLASM_DISEASESTAGE', 'PATHOLOGY_T_STAGE', 'KARNOFSKY_PERFORMANCE_SCORE', and 'RACE'.
CNMF clustering analysis on RPPA data identified 5 subtypes that do not correlate to any clinical features.
Consensus hierarchical clustering analysis on RPPA data identified 4 subtypes that correlate to 'RACE'.
CNMF clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that correlate to 'Time to Death', 'NEOPLASM_DISEASESTAGE', 'PATHOLOGY_T_STAGE', 'PATHOLOGY_N_STAGE', 'KARNOFSKY_PERFORMANCE_SCORE', and 'RACE'.
Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'NEOPLASM_DISEASESTAGE', 'PATHOLOGY_T_STAGE', 'PATHOLOGY_N_STAGE', 'GENDER', 'KARNOFSKY_PERFORMANCE_SCORE', 'NUMBER_OF_LYMPH_NODES', and 'RACE'.
3 subtypes identified in current cancer cohort by 'MIRSEQ CNMF'. These subtypes correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'NEOPLASM_DISEASESTAGE', 'PATHOLOGY_T_STAGE', 'PATHOLOGY_N_STAGE', and 'NUMBER_OF_LYMPH_NODES'.
5 subtypes identified in current cancer cohort by 'MIRSEQ CHIERARCHICAL'. These subtypes correlate to 'YEARS_TO_BIRTH', 'NEOPLASM_DISEASESTAGE', 'PATHOLOGY_T_STAGE', 'PATHOLOGY_N_STAGE', 'GENDER', 'KARNOFSKY_PERFORMANCE_SCORE', and 'RACE'.
4 subtypes identified in current cancer cohort by 'MIRseq Mature CNMF subtypes'. These subtypes correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'NEOPLASM_DISEASESTAGE', 'PATHOLOGY_T_STAGE', 'PATHOLOGY_N_STAGE', 'KARNOFSKY_PERFORMANCE_SCORE', 'NUMBER_OF_LYMPH_NODES', and 'RACE'.
6 subtypes identified in current cancer cohort by 'MIRseq Mature cHierClus subtypes'. These subtypes correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'NEOPLASM_DISEASESTAGE', 'PATHOLOGY_T_STAGE', 'PATHOLOGY_N_STAGE', 'KARNOFSKY_PERFORMANCE_SCORE', 'NUMBER_OF_LYMPH_NODES', and 'RACE'.

Results

Overview of the results

Table 1. Get Full Table Overview of the association between subtypes identified by 10 different clustering approaches and 12 clinical features. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, 52 significant findings detected.


Clinical Features	Statistical Tests	Copy Number Ratio CNMF subtypes	METHLYATION CNMF	RPPA CNMF subtypes	RPPA cHierClus subtypes	RNAseq CNMF subtypes	RNAseq cHierClus subtypes	MIRSEQ CNMF	MIRSEQ CHIERARCHICAL	MIRseq Mature CNMF subtypes	MIRseq Mature cHierClus subtypes
Time to Death	logrank test	0.546 (0.636)	0.0717 (0.154)	0.0799 (0.166)	0.379 (0.478)	0.00626 (0.0203)	0.00852 (0.0269)	0.0413 (0.0972)	0.338 (0.443)	0.00476 (0.0163)	0.0227 (0.058)
YEARS TO BIRTH	Kruskal-Wallis (anova)	0.0651 (0.142)	0.0139 (0.0378)	0.457 (0.554)	0.322 (0.435)	0.195 (0.308)	9.15e-05 (0.000578)	0.0458 (0.106)	0.00037 (0.00171)	5.11e-05 (0.000361)	0.000205 (0.00112)
NEOPLASM DISEASESTAGE	Fisher's exact test	0.14 (0.24)	0.00076 (0.00314)	0.0963 (0.186)	0.107 (0.198)	1e-05 (9.23e-05)	1e-05 (9.23e-05)	1e-05 (9.23e-05)	0.00061 (0.00261)	1e-05 (9.23e-05)	1e-05 (9.23e-05)
PATHOLOGY T STAGE	Fisher's exact test	0.733 (0.799)	0.012 (0.0334)	0.727 (0.799)	0.36 (0.465)	1e-05 (9.23e-05)	1e-05 (9.23e-05)	5e-05 (0.000361)	8e-05 (0.000533)	1e-05 (9.23e-05)	2e-05 (0.000171)
PATHOLOGY N STAGE	Fisher's exact test	0.00914 (0.0279)	0.117 (0.21)	0.31 (0.428)	0.552 (0.636)	0.0146 (0.039)	1e-05 (9.23e-05)	0.00034 (0.00163)	0.00929 (0.0279)	1e-05 (9.23e-05)	0.00011 (0.000629)
PATHOLOGY M STAGE	Fisher's exact test	0.643 (0.721)	0.894 (0.917)	0.249 (0.37)	0.162 (0.269)	0.0824 (0.166)	0.464 (0.557)	0.828 (0.871)	0.133 (0.231)	0.615 (0.703)	0.215 (0.33)
GENDER	Fisher's exact test	0.825 (0.871)	0.204 (0.318)	0.517 (0.614)	0.17 (0.279)	0.098 (0.187)	0.0101 (0.0295)	0.0829 (0.166)	0.0337 (0.0808)	0.317 (0.432)	0.627 (0.71)
KARNOFSKY PERFORMANCE SCORE	Kruskal-Wallis (anova)	0.253 (0.37)	0.0032 (0.012)	0.369 (0.471)	0.393 (0.482)	0.0334 (0.0808)	0.00474 (0.0163)	0.263 (0.379)	0.0108 (0.031)	0.003 (0.0116)	0.000583 (0.00259)
NUMBER PACK YEARS SMOKED	Kruskal-Wallis (anova)	0.744 (0.804)	0.534 (0.628)	0.187 (0.299)	0.156 (0.264)	0.333 (0.443)	0.835 (0.871)	0.663 (0.736)	0.787 (0.843)	0.851 (0.88)	0.386 (0.481)
NUMBER OF LYMPH NODES	Kruskal-Wallis (anova)	0.11 (0.201)	0.0827 (0.166)	0.0885 (0.174)	0.232 (0.352)	0.252 (0.37)	5.19e-06 (9.23e-05)	0.00115 (0.00458)	0.0605 (0.135)	0.000269 (0.00135)	0.0159 (0.0415)
RACE	Fisher's exact test	0.265 (0.379)	0.00027 (0.00135)	0.0607 (0.135)	0.00442 (0.0161)	1e-05 (9.23e-05)	1e-05 (9.23e-05)	0.174 (0.282)	0.00599 (0.02)	0.0001 (6e-04)	5e-05 (0.000361)
ETHNICITY	Fisher's exact test	0.0285 (0.0713)	0.102 (0.19)	0.389 (0.481)	0.124 (0.219)	0.34 (0.443)	0.269 (0.38)	0.28 (0.39)	0.987 (0.987)	0.931 (0.947)	0.963 (0.971)

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

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

Cluster Labels	1	2	3	4
Number of samples	165	132	55	37

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

P value = 0.546 (logrank test), Q value = 0.64

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

	nPatients	nDeath	Duration Range (Median), Month
ALL	383	153	0.1 - 166.0 (15.7)
subtype1	161	63	0.1 - 140.8 (15.0)
subtype2	131	49	0.4 - 166.0 (17.2)
subtype3	55	26	1.6 - 123.8 (15.1)
subtype4	36	15	2.2 - 112.8 (18.3)

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

'Copy Number Ratio CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.0651 (Kruskal-Wallis (anova)), Q value = 0.14

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

	nPatients	Mean (Std.Dev)
ALL	388	67.8 (10.6)
subtype1	165	66.2 (11.0)
subtype2	131	69.2 (9.7)
subtype3	55	69.6 (9.5)
subtype4	37	67.2 (12.6)

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

'Copy Number Ratio CNMF subtypes' versus 'NEOPLASM_DISEASESTAGE'

P value = 0.14 (Fisher's exact test), Q value = 0.24

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

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	4	124	127	129
subtype1	2	59	61	41
subtype2	2	35	37	55
subtype3	0	18	15	22
subtype4	0	12	14	11

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

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.733 (Fisher's exact test), Q value = 0.8

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

nPatients	T0+T1+T2	T3	T4
ALL	121	184	53
subtype1	55	72	26
subtype2	39	63	19
subtype3	16	29	5
subtype4	11	20	3

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

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.00914 (Fisher's exact test), Q value = 0.028

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

nPatients	N0	N1	N2	N3
ALL	224	42	75	8
subtype1	114	15	20	2
subtype2	65	17	34	4
subtype3	25	5	16	1
subtype4	20	5	5	1

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

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.643 (Fisher's exact test), Q value = 0.72

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

nPatients	0	1
ALL	185	10
subtype1	91	5
subtype2	64	3
subtype3	18	2
subtype4	12	0

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

'Copy Number Ratio CNMF subtypes' versus 'GENDER'

P value = 0.825 (Fisher's exact test), Q value = 0.87

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

nPatients	FEMALE	MALE
ALL	101	288
subtype1	40	125
subtype2	34	98
subtype3	16	39
subtype4	11	26

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

'Copy Number Ratio CNMF subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.253 (Kruskal-Wallis (anova)), Q value = 0.37

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

	nPatients	Mean (Std.Dev)
ALL	125	82.7 (14.0)
subtype1	49	81.4 (13.2)
subtype2	43	84.7 (14.9)
subtype3	17	84.1 (10.0)
subtype4	16	80.0 (17.9)

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

'Copy Number Ratio CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.744 (Kruskal-Wallis (anova)), Q value = 0.8

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

	nPatients	Mean (Std.Dev)
ALL	212	38.7 (54.2)
subtype1	86	35.9 (27.2)
subtype2	72	44.9 (86.2)
subtype3	28	37.9 (21.3)
subtype4	26	32.0 (22.8)

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

'Copy Number Ratio CNMF subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.11 (Kruskal-Wallis (anova)), Q value = 0.2

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

	nPatients	Mean (Std.Dev)
ALL	284	2.0 (6.9)
subtype1	112	1.2 (3.4)
subtype2	103	2.8 (10.1)
subtype3	43	2.4 (5.2)
subtype4	26	1.8 (4.6)

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

'Copy Number Ratio CNMF subtypes' versus 'RACE'

P value = 0.265 (Fisher's exact test), Q value = 0.38

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

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	42	21	311
subtype1	26	9	128
subtype2	11	8	104
subtype3	4	3	47
subtype4	1	1	32

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

'Copy Number Ratio CNMF subtypes' versus 'ETHNICITY'

P value = 0.0285 (Fisher's exact test), Q value = 0.071

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

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	9	352
subtype1	1	157
subtype2	6	112
subtype3	0	51
subtype4	2	32

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

Clustering Approach #2: 'METHLYATION CNMF'

Table S14. Description of clustering approach #2: 'METHLYATION CNMF'

Cluster Labels	1	2	3	4
Number of samples	136	107	66	84

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.0717 (logrank test), Q value = 0.15

Table S15. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	387	155	0.1 - 166.0 (15.6)
subtype1	135	61	0.4 - 166.0 (14.4)
subtype2	107	34	0.4 - 130.9 (17.6)
subtype3	65	23	0.1 - 103.5 (15.0)
subtype4	80	37	0.5 - 86.3 (14.4)

Figure S13. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'YEARS_TO_BIRTH'

P value = 0.0139 (Kruskal-Wallis (anova)), Q value = 0.038

Table S16. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	392	67.8 (10.6)
subtype1	136	69.4 (9.8)
subtype2	106	69.0 (9.8)
subtype3	66	65.2 (10.9)
subtype4	84	65.5 (11.9)

Figure S14. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'METHLYATION CNMF' versus 'NEOPLASM_DISEASESTAGE'

P value = 0.00076 (Fisher's exact test), Q value = 0.0031

Table S17. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	4	126	128	130
subtype1	0	29	63	43
subtype2	3	37	27	39
subtype3	1	25	15	24
subtype4	0	35	23	24

Figure S15. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_T_STAGE'

P value = 0.012 (Fisher's exact test), Q value = 0.033

Table S18. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T0+T1+T2	T3	T4
ALL	122	185	54
subtype1	27	80	20
subtype2	38	44	11
subtype3	24	26	11
subtype4	33	35	12

Figure S16. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_N_STAGE'

P value = 0.117 (Fisher's exact test), Q value = 0.21

Table S19. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2	N3
ALL	227	42	76	8
subtype1	82	14	24	4
subtype2	55	15	21	2
subtype3	37	2	20	1
subtype4	53	11	11	1

Figure S17. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_M_STAGE'

P value = 0.894 (Fisher's exact test), Q value = 0.92

Table S20. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	189	10
subtype1	53	4
subtype2	54	2
subtype3	45	2
subtype4	37	2

Figure S18. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'METHLYATION CNMF' versus 'GENDER'

P value = 0.204 (Fisher's exact test), Q value = 0.32

Table S21. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	102	291
subtype1	40	96
subtype2	31	76
subtype3	11	55
subtype4	20	64

Figure S19. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'GENDER'

'METHLYATION CNMF' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.0032 (Kruskal-Wallis (anova)), Q value = 0.012

Table S22. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	126	82.8 (14.0)
subtype1	36	87.5 (10.5)
subtype2	37	85.1 (11.2)
subtype3	21	78.6 (12.0)
subtype4	32	77.5 (18.8)

Figure S20. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

'METHLYATION CNMF' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.534 (Kruskal-Wallis (anova)), Q value = 0.63

Table S23. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	214	38.8 (54.1)
subtype1	78	37.4 (28.2)
subtype2	57	33.8 (25.9)
subtype3	32	31.7 (21.6)
subtype4	47	52.0 (104.0)

Figure S21. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

'METHLYATION CNMF' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.0827 (Kruskal-Wallis (anova)), Q value = 0.17

Table S24. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	284	2.0 (6.9)
subtype1	110	1.3 (2.6)
subtype2	71	1.6 (3.4)
subtype3	41	3.5 (7.1)
subtype4	62	2.8 (12.6)

Figure S22. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

'METHLYATION CNMF' versus 'RACE'

P value = 0.00027 (Fisher's exact test), Q value = 0.0013

Table S25. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	43	22	311
subtype1	5	8	118
subtype2	10	6	86
subtype3	18	2	44
subtype4	10	6	63

Figure S23. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #11: 'RACE'

'METHLYATION CNMF' versus 'ETHNICITY'

P value = 0.102 (Fisher's exact test), Q value = 0.19

Table S26. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	9	354
subtype1	2	124
subtype2	2	94
subtype3	0	63
subtype4	5	73

Figure S24. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #3: 'RPPA CNMF subtypes'

Table S27. Description of clustering approach #3: 'RPPA CNMF subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	37	29	36	5	20

'RPPA CNMF subtypes' versus 'Time to Death'

P value = 0.0799 (logrank test), Q value = 0.17

Table S28. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	122	59	0.1 - 140.8 (17.0)
subtype1	33	9	0.1 - 130.9 (17.9)
subtype2	28	18	1.8 - 140.8 (15.5)
subtype3	36	18	0.4 - 123.8 (14.9)
subtype4	5	3	3.1 - 17.7 (15.9)
subtype5	20	11	7.2 - 61.9 (20.6)

Figure S25. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.457 (Kruskal-Wallis (anova)), Q value = 0.55

Table S29. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	126	67.2 (10.5)
subtype1	36	65.3 (10.7)
subtype2	29	69.4 (9.5)
subtype3	36	68.7 (10.6)
subtype4	5	64.4 (11.4)
subtype5	20	65.0 (11.0)

Figure S26. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RPPA CNMF subtypes' versus 'NEOPLASM_DISEASESTAGE'

P value = 0.0963 (Fisher's exact test), Q value = 0.19

Table S30. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	35	43	44
subtype1	1	11	9	13
subtype2	0	5	8	16
subtype3	0	8	17	10
subtype4	0	1	3	1
subtype5	0	10	6	4

Figure S27. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.727 (Fisher's exact test), Q value = 0.8

Table S31. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T0+T1+T2	T3	T4
ALL	27	63	20
subtype1	10	15	5
subtype2	6	15	8
subtype3	8	21	5
subtype4	0	3	1
subtype5	3	9	1

Figure S28. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.31 (Fisher's exact test), Q value = 0.43

Table S32. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2	N3
ALL	71	13	26	6
subtype1	20	1	10	2
subtype2	13	4	10	2
subtype3	23	5	4	2
subtype4	4	1	0	0
subtype5	11	2	2	0

Figure S29. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.249 (Fisher's exact test), Q value = 0.37

Table S33. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	73	5
subtype1	25	2
subtype2	10	2
subtype3	25	0
subtype4	4	0
subtype5	9	1

Figure S30. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RPPA CNMF subtypes' versus 'GENDER'

P value = 0.517 (Fisher's exact test), Q value = 0.61

Table S34. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	33	94
subtype1	7	30
subtype2	10	19
subtype3	8	28
subtype4	2	3
subtype5	6	14

Figure S31. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'RPPA CNMF subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.369 (Kruskal-Wallis (anova)), Q value = 0.47

Table S35. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	35	79.1 (15.8)
subtype1	12	75.0 (17.8)
subtype2	2	80.0 (14.1)
subtype3	10	77.0 (20.0)
subtype4	1	90.0 (NA)
subtype5	10	85.0 (7.1)

Figure S32. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

'RPPA CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.187 (Kruskal-Wallis (anova)), Q value = 0.3

Table S36. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	79	36.8 (23.8)
subtype1	24	34.7 (28.9)
subtype2	18	44.4 (22.5)
subtype3	22	36.8 (20.3)
subtype4	2	35.5 (6.4)
subtype5	13	30.1 (22.2)

Figure S33. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

'RPPA CNMF subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.0885 (Kruskal-Wallis (anova)), Q value = 0.17

Table S37. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	91	1.8 (3.8)
subtype1	23	1.8 (3.0)
subtype2	26	3.2 (5.4)
subtype3	27	0.9 (2.4)
subtype4	4	0.2 (0.5)
subtype5	11	1.4 (3.6)

Figure S34. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

'RPPA CNMF subtypes' versus 'RACE'

P value = 0.0607 (Fisher's exact test), Q value = 0.13

Table S38. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	8	9	103
subtype1	7	1	27
subtype2	0	2	26
subtype3	1	3	29
subtype4	0	0	5
subtype5	0	3	16

Figure S35. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #11: 'RACE'

'RPPA CNMF subtypes' versus 'ETHNICITY'

P value = 0.389 (Fisher's exact test), Q value = 0.48

Table S39. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	3	114
subtype1	0	33
subtype2	2	26
subtype3	1	31
subtype4	0	5
subtype5	0	19

Figure S36. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #4: 'RPPA cHierClus subtypes'

Table S40. Description of clustering approach #4: 'RPPA cHierClus subtypes'

Cluster Labels	1	2	3	4
Number of samples	43	38	27	19

'RPPA cHierClus subtypes' versus 'Time to Death'

P value = 0.379 (logrank test), Q value = 0.48

Table S41. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	122	59	0.1 - 140.8 (17.0)
subtype1	39	13	0.1 - 112.4 (17.6)
subtype2	37	21	1.8 - 140.8 (16.7)
subtype3	27	15	0.4 - 130.9 (16.1)
subtype4	19	10	7.2 - 61.9 (19.4)

Figure S37. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.322 (Kruskal-Wallis (anova)), Q value = 0.43

Table S42. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	126	67.2 (10.5)
subtype1	42	65.4 (10.8)
subtype2	38	68.7 (10.4)
subtype3	27	69.0 (10.9)
subtype4	19	65.3 (9.0)

Figure S38. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RPPA cHierClus subtypes' versus 'NEOPLASM_DISEASESTAGE'

P value = 0.107 (Fisher's exact test), Q value = 0.2

Table S43. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	35	43	44
subtype1	1	14	12	13
subtype2	0	7	12	19
subtype3	0	5	14	7
subtype4	0	9	5	5

Figure S39. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.36 (Fisher's exact test), Q value = 0.46

Table S44. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T0+T1+T2	T3	T4
ALL	27	63	20
subtype1	14	16	7
subtype2	7	22	8
subtype3	4	18	4
subtype4	2	7	1

Figure S40. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.552 (Fisher's exact test), Q value = 0.64

Table S45. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2	N3
ALL	71	13	26	6
subtype1	24	4	9	1
subtype2	19	6	11	2
subtype3	19	3	4	1
subtype4	9	0	2	2

Figure S41. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.162 (Fisher's exact test), Q value = 0.27

Table S46. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	73	5
subtype1	30	1
subtype2	18	2
subtype3	16	0
subtype4	9	2

Figure S42. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RPPA cHierClus subtypes' versus 'GENDER'

P value = 0.17 (Fisher's exact test), Q value = 0.28

Table S47. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	33	94
subtype1	7	36
subtype2	14	24
subtype3	6	21
subtype4	6	13

Figure S43. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'RPPA cHierClus subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.393 (Kruskal-Wallis (anova)), Q value = 0.48

Table S48. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	35	79.1 (15.8)
subtype1	12	75.8 (17.8)
subtype2	5	74.0 (20.7)
subtype3	8	78.8 (17.3)
subtype4	10	86.0 (7.0)

Figure S44. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

'RPPA cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.156 (Kruskal-Wallis (anova)), Q value = 0.26

Table S49. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	79	36.8 (23.8)
subtype1	27	34.8 (26.9)
subtype2	24	43.2 (24.2)
subtype3	16	35.9 (18.1)
subtype4	12	29.5 (21.8)

Figure S45. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

'RPPA cHierClus subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.232 (Kruskal-Wallis (anova)), Q value = 0.35

Table S50. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	91	1.8 (3.8)
subtype1	28	1.6 (2.8)
subtype2	33	2.6 (4.9)
subtype3	21	0.5 (0.9)
subtype4	9	2.9 (5.2)

Figure S46. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

'RPPA cHierClus subtypes' versus 'RACE'

P value = 0.00442 (Fisher's exact test), Q value = 0.016

Table S51. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	8	9	103
subtype1	7	1	32
subtype2	0	1	35
subtype3	1	4	20
subtype4	0	3	16

Figure S47. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #11: 'RACE'

'RPPA cHierClus subtypes' versus 'ETHNICITY'

P value = 0.124 (Fisher's exact test), Q value = 0.22

Table S52. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	3	114
subtype1	0	38
subtype2	3	33
subtype3	0	24
subtype4	0	19

Figure S48. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #5: 'RNAseq CNMF subtypes'

Table S53. Description of clustering approach #5: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	110	172	107

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.00626 (logrank test), Q value = 0.02

Table S54. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	383	153	0.1 - 166.0 (15.6)
subtype1	109	55	0.7 - 166.0 (18.0)
subtype2	167	45	0.1 - 112.4 (15.0)
subtype3	107	53	0.4 - 125.5 (15.1)

Figure S49. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.195 (Kruskal-Wallis (anova)), Q value = 0.31

Table S55. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	388	67.7 (10.6)
subtype1	110	68.8 (9.1)
subtype2	171	66.6 (11.3)
subtype3	107	68.4 (11.0)

Figure S50. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RNAseq CNMF subtypes' versus 'NEOPLASM_DISEASESTAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S56. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	4	125	127	128
subtype1	0	17	46	47
subtype2	4	76	42	47
subtype3	0	32	39	34

Figure S51. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S57. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T0+T1+T2	T3	T4
ALL	121	183	53
subtype1	17	73	16
subtype2	74	57	22
subtype3	30	53	15

Figure S52. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.0146 (Fisher's exact test), Q value = 0.039

Table S58. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2	N3
ALL	225	41	75	8
subtype1	59	13	30	2
subtype2	105	10	32	3
subtype3	61	18	13	3

Figure S53. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.0824 (Fisher's exact test), Q value = 0.17

Table S59. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	189	10
subtype1	38	5
subtype2	105	3
subtype3	46	2

Figure S54. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 0.098 (Fisher's exact test), Q value = 0.19

Table S60. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	101	288
subtype1	27	83
subtype2	38	134
subtype3	36	71

Figure S55. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'RNAseq CNMF subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.0334 (Kruskal-Wallis (anova)), Q value = 0.081

Table S61. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	124	82.7 (14.1)
subtype1	30	87.3 (9.1)
subtype2	61	80.2 (15.0)
subtype3	33	83.0 (15.3)

Figure S56. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

'RNAseq CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.333 (Kruskal-Wallis (anova)), Q value = 0.44

Table S62. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	212	38.9 (54.3)
subtype1	68	36.4 (28.1)
subtype2	89	34.4 (27.3)
subtype3	55	49.4 (95.8)

Figure S57. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

'RNAseq CNMF subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.252 (Kruskal-Wallis (anova)), Q value = 0.37

Table S63. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	280	2.0 (6.9)
subtype1	94	2.2 (4.5)
subtype2	106	2.5 (10.1)
subtype3	80	1.1 (3.0)

Figure S58. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

'RNAseq CNMF subtypes' versus 'RACE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S64. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	43	22	308
subtype1	3	5	101
subtype2	37	9	120
subtype3	3	8	87

Figure S59. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #11: 'RACE'

'RNAseq CNMF subtypes' versus 'ETHNICITY'

P value = 0.34 (Fisher's exact test), Q value = 0.44

Table S65. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	8	351
subtype1	2	104
subtype2	2	156
subtype3	4	91

Figure S60. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #6: 'RNAseq cHierClus subtypes'

Table S66. Description of clustering approach #6: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	108	109	172

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.00852 (logrank test), Q value = 0.027

Table S67. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	383	153	0.1 - 166.0 (15.6)
subtype1	107	47	0.5 - 130.9 (18.0)
subtype2	104	23	0.1 - 91.6 (13.8)
subtype3	172	83	0.4 - 166.0 (14.7)

Figure S61. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 9.15e-05 (Kruskal-Wallis (anova)), Q value = 0.00058

Table S68. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	388	67.7 (10.6)
subtype1	108	70.9 (9.5)
subtype2	108	64.5 (10.6)
subtype3	172	67.8 (10.8)

Figure S62. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RNAseq cHierClus subtypes' versus 'NEOPLASM_DISEASESTAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S69. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	4	125	127	128
subtype1	0	15	33	59
subtype2	4	63	24	16
subtype3	0	47	70	53

Figure S63. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S70. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T0+T1+T2	T3	T4
ALL	121	183	53
subtype1	20	62	21
subtype2	59	26	10
subtype3	42	95	22

Figure S64. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S71. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2	N3
ALL	225	41	75	8
subtype1	47	14	40	3
subtype2	77	4	10	1
subtype3	101	23	25	4

Figure S65. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.464 (Fisher's exact test), Q value = 0.56

Table S72. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	189	10
subtype1	47	3
subtype2	75	2
subtype3	67	5

Figure S66. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 0.0101 (Fisher's exact test), Q value = 0.03

Table S73. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	101	288
subtype1	27	81
subtype2	18	91
subtype3	56	116

Figure S67. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'RNAseq cHierClus subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.00474 (Kruskal-Wallis (anova)), Q value = 0.016

Table S74. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	124	82.7 (14.1)
subtype1	28	85.0 (12.6)
subtype2	41	77.8 (15.6)
subtype3	55	85.1 (12.9)

Figure S68. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

'RNAseq cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.835 (Kruskal-Wallis (anova)), Q value = 0.87

Table S75. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	212	38.9 (54.3)
subtype1	53	38.6 (32.3)
subtype2	62	34.4 (25.4)
subtype3	97	42.0 (74.0)

Figure S69. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

'RNAseq cHierClus subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 5.19e-06 (Kruskal-Wallis (anova)), Q value = 9.2e-05

Table S76. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	280	2.0 (6.9)
subtype1	90	2.4 (3.6)
subtype2	57	2.9 (13.5)
subtype3	133	1.4 (3.9)

Figure S70. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

'RNAseq cHierClus subtypes' versus 'RACE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S77. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	43	22	308
subtype1	5	8	93
subtype2	32	3	70
subtype3	6	11	145

Figure S71. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #11: 'RACE'

'RNAseq cHierClus subtypes' versus 'ETHNICITY'

P value = 0.269 (Fisher's exact test), Q value = 0.38

Table S78. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	8	351
subtype1	1	97
subtype2	1	102
subtype3	6	152

Figure S72. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #7: 'MIRSEQ CNMF'

Table S79. Description of clustering approach #7: 'MIRSEQ CNMF'

Cluster Labels	1	2	3
Number of samples	116	93	182

'MIRSEQ CNMF' versus 'Time to Death'

P value = 0.0413 (logrank test), Q value = 0.097

Table S80. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	385	154	0.1 - 166.0 (15.7)
subtype1	116	58	0.4 - 166.0 (14.9)
subtype2	90	35	1.2 - 130.9 (15.2)
subtype3	179	61	0.1 - 125.5 (16.7)

Figure S73. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CNMF' versus 'YEARS_TO_BIRTH'

P value = 0.0458 (Kruskal-Wallis (anova)), Q value = 0.11

Table S81. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	390	67.7 (10.6)
subtype1	116	69.1 (9.8)
subtype2	92	68.6 (11.2)
subtype3	182	66.4 (10.7)

Figure S74. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRSEQ CNMF' versus 'NEOPLASM_DISEASESTAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S82. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	4	126	127	129
subtype1	0	22	43	48
subtype2	1	23	29	39
subtype3	3	81	55	42

Figure S75. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_T_STAGE'

P value = 5e-05 (Fisher's exact test), Q value = 0.00036

Table S83. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T0+T1+T2	T3	T4
ALL	122	184	53
subtype1	21	76	13
subtype2	28	45	15
subtype3	73	63	25

Figure S76. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_N_STAGE'

P value = 0.00034 (Fisher's exact test), Q value = 0.0016

Table S84. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2	N3
ALL	226	41	76	8
subtype1	60	13	35	0
subtype2	46	13	21	4
subtype3	120	15	20	4

Figure S77. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_M_STAGE'

P value = 0.828 (Fisher's exact test), Q value = 0.87

Table S85. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	188	10
subtype1	43	2
subtype2	40	3
subtype3	105	5

Figure S78. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRSEQ CNMF' versus 'GENDER'

P value = 0.0829 (Fisher's exact test), Q value = 0.17

Table S86. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	101	290
subtype1	38	78
subtype2	18	75
subtype3	45	137

Figure S79. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #7: 'GENDER'

'MIRSEQ CNMF' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.263 (Kruskal-Wallis (anova)), Q value = 0.38

Table S87. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	126	82.8 (14.0)
subtype1	29	85.2 (12.7)
subtype2	25	83.6 (15.0)
subtype3	72	81.5 (14.2)

Figure S80. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

'MIRSEQ CNMF' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.663 (Kruskal-Wallis (anova)), Q value = 0.74

Table S88. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	212	39.0 (54.3)
subtype1	64	49.0 (91.4)
subtype2	48	35.2 (28.6)
subtype3	100	34.4 (21.8)

Figure S81. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

'MIRSEQ CNMF' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.00115 (Kruskal-Wallis (anova)), Q value = 0.0046

Table S89. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	283	2.0 (6.9)
subtype1	92	2.0 (4.2)
subtype2	73	3.6 (11.9)
subtype3	118	1.1 (3.4)

Figure S82. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

'MIRSEQ CNMF' versus 'RACE'

P value = 0.174 (Fisher's exact test), Q value = 0.28

Table S90. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	43	22	309
subtype1	7	7	101
subtype2	10	4	75
subtype3	26	11	133

Figure S83. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #11: 'RACE'

'MIRSEQ CNMF' versus 'ETHNICITY'

P value = 0.28 (Fisher's exact test), Q value = 0.39

Table S91. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	9	352
subtype1	1	111
subtype2	4	81
subtype3	4	160

Figure S84. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #8: 'MIRSEQ CHIERARCHICAL'

Table S92. Description of clustering approach #8: 'MIRSEQ CHIERARCHICAL'

Cluster Labels	1	2	3	4	5
Number of samples	85	95	94	66	51

'MIRSEQ CHIERARCHICAL' versus 'Time to Death'

P value = 0.338 (logrank test), Q value = 0.44

Table S93. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	385	154	0.1 - 166.0 (15.7)
subtype1	85	41	0.4 - 166.0 (16.1)
subtype2	93	40	1.2 - 130.9 (15.7)
subtype3	90	29	0.1 - 103.5 (16.3)
subtype4	66	23	0.4 - 110.6 (14.5)
subtype5	51	21	0.5 - 140.8 (16.2)

Figure S85. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CHIERARCHICAL' versus 'YEARS_TO_BIRTH'

P value = 0.00037 (Kruskal-Wallis (anova)), Q value = 0.0017

Table S94. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	390	67.7 (10.6)
subtype1	85	67.6 (11.2)
subtype2	94	70.1 (10.1)
subtype3	94	64.1 (10.8)
subtype4	66	67.4 (9.9)
subtype5	51	70.6 (9.6)

Figure S86. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRSEQ CHIERARCHICAL' versus 'NEOPLASM_DISEASESTAGE'

P value = 0.00061 (Fisher's exact test), Q value = 0.0026

Table S95. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	4	126	127	129
subtype1	0	21	35	27
subtype2	1	23	30	40
subtype3	2	47	21	23
subtype4	1	25	24	16
subtype5	0	10	17	23

Figure S87. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_T_STAGE'

P value = 8e-05 (Fisher's exact test), Q value = 0.00053

Table S96. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T0+T1+T2	T3	T4
ALL	122	184	53
subtype1	18	50	9
subtype2	26	47	16
subtype3	49	25	12
subtype4	18	30	9
subtype5	11	32	7

Figure S88. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_N_STAGE'

P value = 0.00929 (Fisher's exact test), Q value = 0.028

Table S97. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2	N3
ALL	226	41	76	8
subtype1	48	14	11	2
subtype2	49	13	23	4
subtype3	58	6	14	1
subtype4	44	5	8	1
subtype5	27	3	20	0

Figure S89. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_M_STAGE'

P value = 0.133 (Fisher's exact test), Q value = 0.23

Table S98. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	188	10
subtype1	38	0
subtype2	40	1
subtype3	60	3
subtype4	33	4
subtype5	17	2

Figure S90. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'GENDER'

P value = 0.0337 (Fisher's exact test), Q value = 0.081

Table S99. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	101	290
subtype1	28	57
subtype2	18	77
subtype3	20	74
subtype4	15	51
subtype5	20	31

Figure S91. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'GENDER'

'MIRSEQ CHIERARCHICAL' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.0108 (Kruskal-Wallis (anova)), Q value = 0.031

Table S100. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	126	82.8 (14.0)
subtype1	25	83.6 (15.2)
subtype2	25	86.4 (9.5)
subtype3	37	76.5 (17.2)
subtype4	26	85.0 (9.9)
subtype5	13	87.7 (10.9)

Figure S92. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

'MIRSEQ CHIERARCHICAL' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.787 (Kruskal-Wallis (anova)), Q value = 0.84

Table S101. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	212	39.0 (54.3)
subtype1	43	53.3 (108.7)
subtype2	52	33.6 (24.2)
subtype3	53	34.4 (22.1)
subtype4	35	33.5 (24.6)
subtype5	29	42.6 (36.1)

Figure S93. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

'MIRSEQ CHIERARCHICAL' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.0605 (Kruskal-Wallis (anova)), Q value = 0.13

Table S102. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	283	2.0 (6.9)
subtype1	63	1.1 (2.3)
subtype2	78	3.4 (11.5)
subtype3	52	1.9 (6.1)
subtype4	45	0.8 (1.7)
subtype5	45	2.2 (3.7)

Figure S94. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

'MIRSEQ CHIERARCHICAL' versus 'RACE'

P value = 0.00599 (Fisher's exact test), Q value = 0.02

Table S103. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	43	22	309
subtype1	4	4	72
subtype2	8	5	80
subtype3	22	5	63
subtype4	7	5	48
subtype5	2	3	46

Figure S95. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #11: 'RACE'

'MIRSEQ CHIERARCHICAL' versus 'ETHNICITY'

P value = 0.987 (Fisher's exact test), Q value = 0.99

Table S104. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	9	352
subtype1	2	77
subtype2	2	86
subtype3	2	84
subtype4	2	56
subtype5	1	49

Figure S96. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #9: 'MIRseq Mature CNMF subtypes'

Table S105. Description of clustering approach #9: 'MIRseq Mature CNMF subtypes'

Cluster Labels	1	2	3	4
Number of samples	96	40	129	115

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

P value = 0.00476 (logrank test), Q value = 0.016

Table S106. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	374	152	0.4 - 166.0 (16.0)
subtype1	96	40	0.4 - 125.5 (15.5)
subtype2	39	24	1.5 - 130.9 (12.8)
subtype3	124	34	0.4 - 110.6 (17.0)
subtype4	115	54	0.5 - 166.0 (16.3)

Figure S97. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 5.11e-05 (Kruskal-Wallis (anova)), Q value = 0.00036

Table S107. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	379	67.8 (10.5)
subtype1	96	67.9 (11.1)
subtype2	39	71.9 (8.5)
subtype3	129	64.6 (10.5)
subtype4	115	70.1 (9.7)

Figure S98. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRseq Mature CNMF subtypes' versus 'NEOPLASM_DISEASESTAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S108. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	4	119	123	129
subtype1	1	29	34	31
subtype2	1	4	11	23
subtype3	2	63	36	27
subtype4	0	23	42	48

Figure S99. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S109. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T0+T1+T2	T3	T4
ALL	118	180	53
subtype1	26	51	10
subtype2	9	21	10
subtype3	59	37	18
subtype4	24	71	15

Figure S100. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S110. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2	N3
ALL	219	41	76	8
subtype1	57	15	11	3
subtype2	14	6	14	3
subtype3	87	8	14	2
subtype4	61	12	37	0

Figure S101. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.615 (Fisher's exact test), Q value = 0.7

Table S111. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	181	10
subtype1	36	2
subtype2	14	1
subtype3	83	3
subtype4	48	4

Figure S102. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRseq Mature CNMF subtypes' versus 'GENDER'

P value = 0.317 (Fisher's exact test), Q value = 0.43

Table S112. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	101	279
subtype1	28	68
subtype2	9	31
subtype3	28	101
subtype4	36	79

Figure S103. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'MIRseq Mature CNMF subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.003 (Kruskal-Wallis (anova)), Q value = 0.012

Table S113. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	121	82.8 (14.2)
subtype1	28	84.6 (13.7)
subtype2	11	85.5 (12.1)
subtype3	47	77.4 (16.5)
subtype4	35	87.7 (9.1)

Figure S104. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

'MIRseq Mature CNMF subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.851 (Kruskal-Wallis (anova)), Q value = 0.88

Table S114. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	207	39.4 (54.7)
subtype1	53	37.2 (23.2)
subtype2	26	37.7 (26.0)
subtype3	70	34.6 (24.3)
subtype4	58	48.1 (96.0)

Figure S105. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

'MIRseq Mature CNMF subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.000269 (Kruskal-Wallis (anova)), Q value = 0.0013

Table S115. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	279	2.1 (7.0)
subtype1	72	0.9 (1.9)
subtype2	35	3.6 (5.7)
subtype3	78	2.3 (11.5)
subtype4	94	2.2 (4.2)

Figure S106. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

'MIRseq Mature CNMF subtypes' versus 'RACE'

P value = 1e-04 (Fisher's exact test), Q value = 6e-04

Table S116. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	40	21	302
subtype1	3	7	80
subtype2	1	2	36
subtype3	28	6	85
subtype4	8	6	101

Figure S107. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #11: 'RACE'

'MIRseq Mature CNMF subtypes' versus 'ETHNICITY'

P value = 0.931 (Fisher's exact test), Q value = 0.95

Table S117. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	9	341
subtype1	3	84
subtype2	0	37
subtype3	3	111
subtype4	3	109

Figure S108. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #10: 'MIRseq Mature cHierClus subtypes'

Table S118. Description of clustering approach #10: 'MIRseq Mature cHierClus subtypes'

Cluster Labels	1	2	3	4	5	6
Number of samples	71	43	92	63	30	81

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

P value = 0.0227 (logrank test), Q value = 0.058

Table S119. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	374	152	0.4 - 166.0 (16.0)
subtype1	71	30	0.5 - 140.8 (15.8)
subtype2	42	22	1.5 - 130.9 (13.0)
subtype3	89	23	0.4 - 110.6 (17.2)
subtype4	63	36	1.8 - 166.0 (17.7)
subtype5	28	5	0.7 - 82.5 (15.5)
subtype6	81	36	0.4 - 125.5 (15.9)

Figure S109. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.000205 (Kruskal-Wallis (anova)), Q value = 0.0011

Table S120. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	379	67.8 (10.5)
subtype1	71	69.0 (9.8)
subtype2	42	71.1 (10.0)
subtype3	92	64.7 (10.1)
subtype4	63	70.7 (8.9)
subtype5	30	63.2 (10.6)
subtype6	81	68.2 (11.7)

Figure S110. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRseq Mature cHierClus subtypes' versus 'NEOPLASM_DISEASESTAGE'

P value = 1e-05 (Fisher's exact test), Q value = 9.2e-05

Table S121. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	4	119	123	129
subtype1	0	18	25	27
subtype2	1	8	12	21
subtype3	2	40	23	27
subtype4	0	12	25	25
subtype5	1	20	6	2
subtype6	0	21	32	27

Figure S111. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'NEOPLASM_DISEASESTAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 2e-05 (Fisher's exact test), Q value = 0.00017

Table S122. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T0+T1+T2	T3	T4
ALL	118	180	53
subtype1	17	42	8
subtype2	11	22	9
subtype3	36	28	11
subtype4	13	38	10
subtype5	21	4	4
subtype6	20	46	11

Figure S112. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.00011 (Fisher's exact test), Q value = 0.00063

Table S123. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	N0	N1	N2	N3
ALL	219	41	76	8
subtype1	39	4	23	0
subtype2	18	6	13	3
subtype3	53	6	14	2
subtype4	36	11	15	0
subtype5	25	0	1	0
subtype6	48	14	10	3

Figure S113. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.215 (Fisher's exact test), Q value = 0.33

Table S124. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	181	10
subtype1	30	2
subtype2	19	0
subtype3	51	6
subtype4	26	2
subtype5	25	0
subtype6	30	0

Figure S114. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'GENDER'

P value = 0.627 (Fisher's exact test), Q value = 0.71

Table S125. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	101	279
subtype1	23	48
subtype2	10	33
subtype3	22	70
subtype4	18	45
subtype5	5	25
subtype6	23	58

Figure S115. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'MIRseq Mature cHierClus subtypes' versus 'KARNOFSKY_PERFORMANCE_SCORE'

P value = 0.000583 (Kruskal-Wallis (anova)), Q value = 0.0026

Table S126. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

	nPatients	Mean (Std.Dev)
ALL	121	82.8 (14.2)
subtype1	19	86.3 (10.1)
subtype2	11	84.5 (12.1)
subtype3	36	81.7 (15.6)
subtype4	19	88.9 (7.4)
subtype5	11	66.4 (16.3)
subtype6	25	83.6 (14.1)

Figure S116. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'KARNOFSKY_PERFORMANCE_SCORE'

'MIRseq Mature cHierClus subtypes' versus 'NUMBER_PACK_YEARS_SMOKED'

P value = 0.386 (Kruskal-Wallis (anova)), Q value = 0.48

Table S127. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

	nPatients	Mean (Std.Dev)
ALL	207	39.4 (54.7)
subtype1	37	38.3 (34.3)
subtype2	24	33.4 (22.5)
subtype3	50	34.4 (29.6)
subtype4	33	60.2 (122.0)
subtype5	17	39.9 (19.8)
subtype6	46	34.0 (21.3)

Figure S117. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #9: 'NUMBER_PACK_YEARS_SMOKED'

'MIRseq Mature cHierClus subtypes' versus 'NUMBER_OF_LYMPH_NODES'

P value = 0.0159 (Kruskal-Wallis (anova)), Q value = 0.042

Table S128. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

	nPatients	Mean (Std.Dev)
ALL	279	2.1 (7.0)
subtype1	54	1.9 (3.5)
subtype2	34	3.7 (5.9)
subtype3	58	2.6 (12.8)
subtype4	55	1.3 (1.9)
subtype5	13	2.5 (8.9)
subtype6	65	1.4 (4.1)

Figure S118. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'NUMBER_OF_LYMPH_NODES'

'MIRseq Mature cHierClus subtypes' versus 'RACE'

P value = 5e-05 (Fisher's exact test), Q value = 0.00036

Table S129. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	40	21	302
subtype1	5	3	63
subtype2	3	2	36
subtype3	13	6	66
subtype4	2	4	53
subtype5	13	2	14
subtype6	4	4	70

Figure S119. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #11: 'RACE'

'MIRseq Mature cHierClus subtypes' versus 'ETHNICITY'

P value = 0.963 (Fisher's exact test), Q value = 0.97

Table S130. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	9	341
subtype1	1	68
subtype2	1	37
subtype3	3	77
subtype4	2	57
subtype5	0	28
subtype6	2	74

Figure S120. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

Methods & Data

Input

Cluster data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/GDAC_mergedClustering/BLCA-TP/15111049/BLCA-TP.mergedcluster.txt
Clinical data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/Append_Data/BLCA-TP/15076720/BLCA-TP.merged_data.txt
Number of patients = 393
Number of clustering approaches = 10
Number of selected clinical features = 12
Exclude small clusters that include fewer than K patients, K = 3

Clustering approaches

CNMF clustering

consensus non-negative matrix factorization clustering approach (Brunet et al. 2004)

Consensus hierarchical clustering

Resampling-based clustering method (Monti et al. 2003)

Survival analysis

For survival clinical features, the Kaplan-Meier survival curves of tumors with and without gene mutations were plotted and the statistical significance P values were estimated by logrank test (Bland and Altman 2004) using the 'survdiff' function in R

Fisher's exact test

For binary clinical features, two-tailed Fisher's exact tests (Fisher 1922) were used to estimate the P values using the 'fisher.test' function in R

Q value calculation

For multiple hypothesis correction, Q value is the False Discovery Rate (FDR) analogue of the P value (Benjamini and Hochberg 1995), defined as the minimum FDR at which the test may be called significant. We used the 'Benjamini and Hochberg' method of 'p.adjust' function in R to convert P values into Q values.

References

[1] Brunet et al., Metagenes and molecular pattern discovery using matrix factorization, PNAS 101(12):4164-9 (2004)

[2] Monti et al., Consensus Clustering: A Resampling-Based Method for Class Discovery and Visualization of Gene Expression Microarray Data, Machine Learning 52(1):91-118 (2003)

[3] Bland and Altman, Statistics notes: The logrank test, BMJ 328(7447):1073 (2004)

[4] Fisher, R.A., On the interpretation of chi-square from contingency tables, and the calculation of P, Journal of the Royal Statistical Society 85(1):87-94 (1922)

[5] Benjamini and Hochberg, Controlling the false discovery rate: a practical and powerful approach to multiple testing, Journal of the Royal Statistical Society Series B 59:289-300 (1995)

Made with Nozzle