Correlation between aggregated molecular cancer subtypes and selected clinical features

Bladder Urothelial Carcinoma (Primary solid tumor)

23 May 2013 | analyses__2013_05_23

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 (2013): Correlation between aggregated molecular cancer subtypes and selected clinical features. Broad Institute of MIT and Harvard. doi:10.7908/C1707ZF9

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 10 clinical features across 137 patients, 3 significant findings detected with P value < 0.05 and Q value < 0.25.

3 subtypes identified in current cancer cohort by 'Copy Number Ratio CNMF subtypes'. These subtypes correlate to 'LYMPH.NODE.METASTASIS'.
4 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. These subtypes do not correlate to any clinical features.
CNMF clustering analysis on RPPA data identified 4 subtypes that do not correlate to any clinical features.
Consensus hierarchical clustering analysis on RPPA data identified 3 subtypes that do not correlate to any clinical features.
CNMF clustering analysis on sequencing-based mRNA expression data identified 4 subtypes that do not correlate to any clinical features.
Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that correlate to 'LYMPH.NODE.METASTASIS'.
3 subtypes identified in current cancer cohort by 'MIRSEQ CNMF'. These subtypes do not correlate to any clinical features.
5 subtypes identified in current cancer cohort by 'MIRSEQ CHIERARCHICAL'. These subtypes do not correlate to any clinical features.
3 subtypes identified in current cancer cohort by 'MIRseq Mature CNMF subtypes'. These subtypes do not correlate to any clinical features.
3 subtypes identified in current cancer cohort by 'MIRseq Mature cHierClus subtypes'. These subtypes correlate to 'LYMPH.NODE.METASTASIS'.

Results

Overview of the results

Table 1. Get Full Table Overview of the association between subtypes identified by 10 different clustering approaches and 10 clinical features. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, 3 significant findings detected.


Clinical Features	Time to Death	AGE	GENDER	KARNOFSKY PERFORMANCE SCORE	NUMBERPACKYEARSSMOKED	DISTANT METASTASIS	LYMPH NODE METASTASIS	NUMBER OF LYMPH NODES	TUMOR STAGECODE	NEOPLASM DISEASESTAGE
Statistical Tests	logrank test	ANOVA	Fisher's exact test	ANOVA	ANOVA	Chi-square test	Chi-square test	ANOVA	ANOVA	Chi-square test
Copy Number Ratio CNMF subtypes	0.983 (1.00)	0.306 (1.00)	1 (1.00)	0.594 (1.00)	0.0203 (1.00)	0.538 (1.00)	0.000132 (0.0118)	0.461 (1.00)		0.0296 (1.00)
METHLYATION CNMF	0.476 (1.00)	0.115 (1.00)	0.0406 (1.00)	0.821 (1.00)	0.963 (1.00)	0.572 (1.00)	0.159 (1.00)	0.223 (1.00)		0.114 (1.00)
RPPA CNMF subtypes	0.245 (1.00)	0.473 (1.00)	0.633 (1.00)	0.465 (1.00)	0.657 (1.00)	0.108 (1.00)	0.082 (1.00)	0.201 (1.00)		0.279 (1.00)
RPPA cHierClus subtypes	0.485 (1.00)	0.227 (1.00)	0.166 (1.00)	0.359 (1.00)	0.943 (1.00)	0.142 (1.00)	0.556 (1.00)	0.697 (1.00)		0.512 (1.00)
RNAseq CNMF subtypes	0.567 (1.00)	0.47 (1.00)	0.822 (1.00)	0.631 (1.00)	0.71 (1.00)	0.272 (1.00)	0.361 (1.00)	0.278 (1.00)		0.703 (1.00)
RNAseq cHierClus subtypes	0.537 (1.00)	0.414 (1.00)	0.366 (1.00)	0.84 (1.00)	0.318 (1.00)	0.7 (1.00)	1.71e-05 (0.00154)	0.0176 (1.00)		0.00329 (0.286)
MIRSEQ CNMF	0.266 (1.00)	0.15 (1.00)	0.0685 (1.00)	0.709 (1.00)	0.388 (1.00)	0.997 (1.00)	0.0185 (1.00)	0.303 (1.00)		0.113 (1.00)
MIRSEQ CHIERARCHICAL	0.612 (1.00)	0.179 (1.00)	0.0441 (1.00)	0.835 (1.00)	0.247 (1.00)	0.914 (1.00)	0.0458 (1.00)	0.016 (1.00)		0.178 (1.00)
MIRseq Mature CNMF subtypes	0.929 (1.00)	0.187 (1.00)	0.0101 (0.859)	0.336 (1.00)	1 (1.00)	0.938 (1.00)	0.00713 (0.613)	0.112 (1.00)		0.05 (1.00)
MIRseq Mature cHierClus subtypes	0.115 (1.00)	0.282 (1.00)	0.1 (1.00)	0.54 (1.00)	0.917 (1.00)	0.953 (1.00)	0.00213 (0.187)	0.0634 (1.00)		0.0221 (1.00)

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

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

Cluster Labels	1	2	3
Number of samples	42	58	34

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

P value = 0.983 (logrank test), Q value = 1

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

	nPatients	nDeath	Duration Range (Median), Month
ALL	127	34	0.1 - 131.2 (7.0)
subtype1	40	12	0.8 - 100.5 (6.6)
subtype2	54	13	0.1 - 131.2 (7.0)
subtype3	33	9	0.4 - 123.8 (7.7)

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 'AGE'

P value = 0.306 (ANOVA), Q value = 1

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

	nPatients	Mean (Std.Dev)
ALL	133	67.5 (10.6)
subtype1	41	69.2 (10.9)
subtype2	58	66.0 (11.4)
subtype3	34	68.1 (8.7)

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

'Copy Number Ratio CNMF subtypes' versus 'GENDER'

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

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

nPatients	FEMALE	MALE
ALL	34	100
subtype1	11	31
subtype2	15	43
subtype3	8	26

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

'Copy Number Ratio CNMF subtypes' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.594 (ANOVA), Q value = 1

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

	nPatients	Mean (Std.Dev)
ALL	38	77.9 (16.1)
subtype1	14	81.4 (12.3)
subtype2	13	75.4 (20.7)
subtype3	11	76.4 (15.0)

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

'Copy Number Ratio CNMF subtypes' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.0203 (ANOVA), Q value = 1

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

	nPatients	Mean (Std.Dev)
ALL	84	36.3 (22.7)
subtype1	27	31.3 (22.9)
subtype2	32	32.4 (18.1)
subtype3	25	46.8 (25.1)

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

'Copy Number Ratio CNMF subtypes' versus 'DISTANT.METASTASIS'

P value = 0.538 (Chi-square test), Q value = 1

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

nPatients	M0	M1	MX
ALL	77	5	51
subtype1	20	2	19
subtype2	37	1	20
subtype3	20	2	12

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

'Copy Number Ratio CNMF subtypes' versus 'LYMPH.NODE.METASTASIS'

P value = 0.000132 (Chi-square test), Q value = 0.012

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

nPatients	N0	N1	N2	N3	NX
ALL	75	14	27	5	10
subtype1	12	7	12	2	7
subtype2	44	1	10	0	2
subtype3	19	6	5	3	1

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

'Copy Number Ratio CNMF subtypes' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.461 (ANOVA), Q value = 1

Table S9. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	100	1.9 (3.7)
subtype1	31	1.9 (2.8)
subtype2	42	1.4 (2.7)
subtype3	27	2.6 (5.6)

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

'Copy Number Ratio CNMF subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.0296 (Chi-square test), Q value = 1

Table S10. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	36	44	48
subtype1	1	11	8	21
subtype2	0	15	27	13
subtype3	0	10	9	14

Figure S9. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #2: 'METHLYATION CNMF'

Table S11. Get Full Table Description of clustering approach #2: 'METHLYATION CNMF'

Cluster Labels	1	2	3	4
Number of samples	36	45	34	22

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.476 (logrank test), Q value = 1

Table S12. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	130	35	0.1 - 131.2 (7.0)
subtype1	35	6	0.8 - 118.9 (6.7)
subtype2	44	15	0.4 - 131.2 (7.8)
subtype3	29	7	0.7 - 37.8 (6.6)
subtype4	22	7	0.1 - 46.8 (6.8)

Figure S10. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'AGE'

P value = 0.115 (ANOVA), Q value = 1

Table S13. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	136	67.6 (10.6)
subtype1	35	68.3 (10.0)
subtype2	45	68.4 (9.6)
subtype3	34	63.9 (12.5)
subtype4	22	70.3 (9.2)

Figure S11. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

'METHLYATION CNMF' versus 'GENDER'

P value = 0.0406 (Fisher's exact test), Q value = 1

Table S14. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	35	102
subtype1	12	24
subtype2	15	30
subtype3	3	31
subtype4	5	17

Figure S12. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'GENDER'

'METHLYATION CNMF' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.821 (ANOVA), Q value = 1

Table S15. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

	nPatients	Mean (Std.Dev)
ALL	39	78.2 (16.0)
subtype1	7	82.9 (15.0)
subtype2	10	79.0 (15.2)
subtype3	17	75.9 (18.4)
subtype4	5	78.0 (13.0)

Figure S13. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

'METHLYATION CNMF' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.963 (ANOVA), Q value = 1

Table S16. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	85	36.8 (22.9)
subtype1	24	37.8 (28.6)
subtype2	31	35.4 (20.0)
subtype3	22	38.3 (21.2)
subtype4	8	35.2 (23.5)

Figure S14. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

'METHLYATION CNMF' versus 'DISTANT.METASTASIS'

P value = 0.572 (Chi-square test), Q value = 1

Table S17. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'DISTANT.METASTASIS'

nPatients	M0	M1	MX
ALL	80	5	51
subtype1	20	2	13
subtype2	25	1	19
subtype3	21	0	13
subtype4	14	2	6

Figure S15. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'DISTANT.METASTASIS'

'METHLYATION CNMF' versus 'LYMPH.NODE.METASTASIS'

P value = 0.159 (Chi-square test), Q value = 1

Table S18. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N2	N3	NX
ALL	77	14	28	5	10
subtype1	14	5	11	1	4
subtype2	31	5	5	2	2
subtype3	24	3	4	1	2
subtype4	8	1	8	1	2

Figure S16. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

'METHLYATION CNMF' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.223 (ANOVA), Q value = 1

Table S19. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	100	1.9 (3.7)
subtype1	26	2.1 (3.4)
subtype2	35	1.1 (2.5)
subtype3	25	1.7 (4.6)
subtype4	14	3.6 (4.8)

Figure S17. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

'METHLYATION CNMF' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.114 (Chi-square test), Q value = 1

Table S20. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	37	45	49
subtype1	1	8	9	17
subtype2	0	10	21	12
subtype3	0	13	11	9
subtype4	0	6	4	11

Figure S18. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #3: 'RPPA CNMF subtypes'

Table S21. Get Full Table Description of clustering approach #3: 'RPPA CNMF subtypes'

Cluster Labels	1	2	3	4
Number of samples	12	10	14	17

'RPPA CNMF subtypes' versus 'Time to Death'

P value = 0.245 (logrank test), Q value = 1

Table S22. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	52	20	0.4 - 118.9 (8.3)
subtype1	12	4	5.7 - 49.9 (10.9)
subtype2	10	5	1.8 - 100.5 (5.5)
subtype3	13	2	0.5 - 19.5 (6.6)
subtype4	17	9	0.4 - 118.9 (14.9)

Figure S19. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA CNMF subtypes' versus 'AGE'

P value = 0.473 (ANOVA), Q value = 1

Table S23. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	52	67.3 (9.8)
subtype1	12	63.9 (9.2)
subtype2	10	69.5 (9.3)
subtype3	13	66.5 (10.2)
subtype4	17	69.1 (10.3)

Figure S20. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'AGE'

'RPPA CNMF subtypes' versus 'GENDER'

P value = 0.633 (Fisher's exact test), Q value = 1

Table S24. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	21	32
subtype1	6	6
subtype2	5	5
subtype3	5	9
subtype4	5	12

Figure S21. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'GENDER'

'RPPA CNMF subtypes' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.465 (ANOVA), Q value = 1

Table S25. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

	nPatients	Mean (Std.Dev)
ALL	11	80.0 (16.7)
subtype1	5	88.0 (4.5)
subtype2	1	70.0 (NA)
subtype3	1	60.0 (NA)
subtype4	4	77.5 (25.0)

Figure S22. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

'RPPA CNMF subtypes' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.657 (ANOVA), Q value = 1

Table S26. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	27	32.1 (17.5)
subtype1	9	27.6 (12.4)
subtype2	5	28.8 (18.4)
subtype3	6	35.2 (20.4)
subtype4	7	37.9 (21.6)

Figure S23. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

'RPPA CNMF subtypes' versus 'DISTANT.METASTASIS'

P value = 0.108 (Chi-square test), Q value = 1

Table S27. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

nPatients	M0	M1	MX
ALL	41	2	10
subtype1	7	2	3
subtype2	10	0	0
subtype3	11	0	3
subtype4	13	0	4

Figure S24. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

'RPPA CNMF subtypes' versus 'LYMPH.NODE.METASTASIS'

P value = 0.082 (Chi-square test), Q value = 1

Table S28. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N2	N3	NX
ALL	33	5	11	1	2
subtype1	9	1	0	0	2
subtype2	4	1	5	0	0
subtype3	10	0	3	0	0
subtype4	10	3	3	1	0

Figure S25. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

'RPPA CNMF subtypes' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.201 (ANOVA), Q value = 1

Table S29. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	34	1.5 (3.0)
subtype1	5	0.0 (0.0)
subtype2	7	3.4 (4.1)
subtype3	9	0.9 (1.5)
subtype4	13	1.5 (3.3)

Figure S26. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

'RPPA CNMF subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.279 (Chi-square test), Q value = 1

Table S30. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	11	19	18
subtype1	1	5	4	2
subtype2	0	2	2	6
subtype3	0	2	6	3
subtype4	0	2	7	7

Figure S27. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #4: 'RPPA cHierClus subtypes'

Table S31. Get Full Table Description of clustering approach #4: 'RPPA cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	14	15	24

'RPPA cHierClus subtypes' versus 'Time to Death'

P value = 0.485 (logrank test), Q value = 1

Table S32. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	52	20	0.4 - 118.9 (8.3)
subtype1	14	9	0.4 - 118.9 (10.4)
subtype2	15	5	1.9 - 49.9 (8.3)
subtype3	23	6	0.5 - 100.5 (6.6)

Figure S28. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA cHierClus subtypes' versus 'AGE'

P value = 0.227 (ANOVA), Q value = 1

Table S33. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	52	67.3 (9.8)
subtype1	14	70.5 (9.0)
subtype2	15	64.2 (9.6)
subtype3	23	67.4 (10.2)

Figure S29. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'RPPA cHierClus subtypes' versus 'GENDER'

P value = 0.166 (Fisher's exact test), Q value = 1

Table S34. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	21	32
subtype1	5	9
subtype2	9	6
subtype3	7	17

Figure S30. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

'RPPA cHierClus subtypes' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.359 (ANOVA), Q value = 1

Table S35. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

	nPatients	Mean (Std.Dev)
ALL	11	80.0 (16.7)
subtype1	3	73.3 (28.9)
subtype2	4	90.0 (0.0)
subtype3	4	75.0 (12.9)

Figure S31. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

'RPPA cHierClus subtypes' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.943 (ANOVA), Q value = 1

Table S36. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	27	32.1 (17.5)
subtype1	5	33.0 (23.3)
subtype2	9	30.4 (15.6)
subtype3	13	33.0 (17.8)

Figure S32. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

'RPPA cHierClus subtypes' versus 'DISTANT.METASTASIS'

P value = 0.142 (Chi-square test), Q value = 1

Table S37. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

nPatients	M0	M1	MX
ALL	41	2	10
subtype1	10	0	4
subtype2	10	2	3
subtype3	21	0	3

Figure S33. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

'RPPA cHierClus subtypes' versus 'LYMPH.NODE.METASTASIS'

P value = 0.556 (Chi-square test), Q value = 1

Table S38. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N2	N3	NX
ALL	33	5	11	1	2
subtype1	8	2	3	1	0
subtype2	11	1	1	0	1
subtype3	14	2	7	0	1

Figure S34. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

'RPPA cHierClus subtypes' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.697 (ANOVA), Q value = 1

Table S39. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	34	1.5 (3.0)
subtype1	10	1.9 (3.7)
subtype2	8	0.8 (1.5)
subtype3	16	1.7 (3.1)

Figure S35. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

'RPPA cHierClus subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.512 (Chi-square test), Q value = 1

Table S40. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	11	19	18
subtype1	0	2	5	6
subtype2	0	6	5	4
subtype3	1	3	9	8

Figure S36. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #5: 'RNAseq CNMF subtypes'

Table S41. Get Full Table Description of clustering approach #5: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3	4
Number of samples	47	22	41	24

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.567 (logrank test), Q value = 1

Table S42. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	127	34	0.1 - 131.2 (7.0)
subtype1	42	8	0.1 - 100.5 (6.6)
subtype2	22	8	0.8 - 118.9 (6.6)
subtype3	40	12	0.4 - 75.3 (7.8)
subtype4	23	6	1.3 - 131.2 (6.8)

Figure S37. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'AGE'

P value = 0.47 (ANOVA), Q value = 1

Table S43. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	133	67.4 (10.6)
subtype1	46	67.2 (11.7)
subtype2	22	65.8 (10.0)
subtype3	41	66.8 (10.6)
subtype4	24	70.4 (8.9)

Figure S38. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 0.822 (Fisher's exact test), Q value = 1

Table S44. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	35	99
subtype1	11	36
subtype2	5	17
subtype3	13	28
subtype4	6	18

Figure S39. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

'RNAseq CNMF subtypes' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.631 (ANOVA), Q value = 1

Table S45. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

	nPatients	Mean (Std.Dev)
ALL	39	78.2 (16.0)
subtype1	17	77.1 (17.2)
subtype2	8	83.8 (9.2)
subtype3	12	78.3 (18.5)
subtype4	2	65.0 (7.1)

Figure S40. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

'RNAseq CNMF subtypes' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.71 (ANOVA), Q value = 1

Table S46. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	83	36.3 (22.8)
subtype1	24	40.4 (30.1)
subtype2	19	35.3 (20.5)
subtype3	28	33.0 (17.9)
subtype4	12	37.2 (20.9)

Figure S41. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

'RNAseq CNMF subtypes' versus 'DISTANT.METASTASIS'

P value = 0.272 (Chi-square test), Q value = 1

Table S47. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

nPatients	M0	M1	MX
ALL	77	5	51
subtype1	31	1	15
subtype2	9	2	10
subtype3	24	0	17
subtype4	13	2	9

Figure S42. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

'RNAseq CNMF subtypes' versus 'LYMPH.NODE.METASTASIS'

P value = 0.361 (Chi-square test), Q value = 1

Table S48. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N2	N3	NX
ALL	75	14	27	5	10
subtype1	25	4	12	1	4
subtype2	12	2	3	1	3
subtype3	26	6	3	2	3
subtype4	12	2	9	1	0

Figure S43. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

'RNAseq CNMF subtypes' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.278 (ANOVA), Q value = 1

Table S49. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	99	1.9 (3.7)
subtype1	30	2.0 (3.2)
subtype2	15	0.7 (1.0)
subtype3	32	1.5 (4.3)
subtype4	22	3.0 (4.5)

Figure S44. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

'RNAseq CNMF subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.703 (Chi-square test), Q value = 1

Table S50. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	37	43	48
subtype1	1	15	12	16
subtype2	0	6	9	7
subtype3	0	11	16	13
subtype4	0	5	6	12

Figure S45. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #6: 'RNAseq cHierClus subtypes'

Table S51. Get Full Table Description of clustering approach #6: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	29	47	58

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.537 (logrank test), Q value = 1

Table S52. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	127	34	0.1 - 131.2 (7.0)
subtype1	29	9	0.5 - 100.5 (5.8)
subtype2	41	7	0.1 - 41.0 (6.6)
subtype3	57	18	0.4 - 131.2 (7.8)

Figure S46. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'AGE'

P value = 0.414 (ANOVA), Q value = 1

Table S53. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	133	67.4 (10.6)
subtype1	29	69.7 (9.5)
subtype2	46	66.3 (11.3)
subtype3	58	67.2 (10.6)

Figure S47. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 0.366 (Fisher's exact test), Q value = 1

Table S54. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	35	99
subtype1	8	21
subtype2	9	38
subtype3	18	40

Figure S48. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

'RNAseq cHierClus subtypes' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.84 (ANOVA), Q value = 1

Table S55. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

	nPatients	Mean (Std.Dev)
ALL	39	78.2 (16.0)
subtype1	6	81.7 (13.3)
subtype2	17	77.1 (17.2)
subtype3	16	78.1 (16.4)

Figure S49. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

'RNAseq cHierClus subtypes' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.318 (ANOVA), Q value = 1

Table S56. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	83	36.3 (22.8)
subtype1	13	44.8 (30.1)
subtype2	29	36.1 (24.7)
subtype3	41	33.7 (18.4)

Figure S50. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

'RNAseq cHierClus subtypes' versus 'DISTANT.METASTASIS'

P value = 0.7 (Chi-square test), Q value = 1

Table S57. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

nPatients	M0	M1	MX
ALL	77	5	51
subtype1	17	2	10
subtype2	28	2	16
subtype3	32	1	25

Figure S51. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

'RNAseq cHierClus subtypes' versus 'LYMPH.NODE.METASTASIS'

P value = 1.71e-05 (Chi-square test), Q value = 0.0015

Table S58. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N2	N3	NX
ALL	75	14	27	5	10
subtype1	10	3	16	0	0
subtype2	25	5	7	1	7
subtype3	40	6	4	4	3

Figure S52. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

'RNAseq cHierClus subtypes' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.0176 (ANOVA), Q value = 1

Table S59. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	99	1.9 (3.7)
subtype1	23	3.8 (4.3)
subtype2	29	1.5 (3.0)
subtype3	47	1.2 (3.6)

Figure S53. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

'RNAseq cHierClus subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.00329 (Chi-square test), Q value = 0.29

Table S60. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	37	43	48
subtype1	0	5	3	19
subtype2	1	16	15	13
subtype3	0	16	25	16

Figure S54. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #7: 'MIRSEQ CNMF'

Table S61. Get Full Table Description of clustering approach #7: 'MIRSEQ CNMF'

Cluster Labels	1	2	3
Number of samples	32	57	48

'MIRSEQ CNMF' versus 'Time to Death'

P value = 0.266 (logrank test), Q value = 1

Table S62. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	130	35	0.1 - 131.2 (7.0)
subtype1	28	3	0.8 - 118.9 (5.4)
subtype2	54	14	0.4 - 49.9 (7.0)
subtype3	48	18	0.1 - 131.2 (8.2)

Figure S55. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CNMF' versus 'AGE'

P value = 0.15 (ANOVA), Q value = 1

Table S63. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	136	67.6 (10.6)
subtype1	31	69.8 (11.3)
subtype2	57	65.6 (10.8)
subtype3	48	68.5 (9.6)

Figure S56. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'AGE'

'MIRSEQ CNMF' versus 'GENDER'

P value = 0.0685 (Fisher's exact test), Q value = 1

Table S64. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	35	102
subtype1	4	28
subtype2	14	43
subtype3	17	31

Figure S57. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'GENDER'

'MIRSEQ CNMF' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.709 (ANOVA), Q value = 1

Table S65. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

	nPatients	Mean (Std.Dev)
ALL	39	78.2 (16.0)
subtype1	7	82.9 (9.5)
subtype2	25	77.2 (18.6)
subtype3	7	77.1 (11.1)

Figure S58. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

'MIRSEQ CNMF' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.388 (ANOVA), Q value = 1

Table S66. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	85	36.8 (22.9)
subtype1	20	37.5 (29.2)
subtype2	40	33.5 (17.3)
subtype3	25	41.5 (25.3)

Figure S59. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

'MIRSEQ CNMF' versus 'DISTANT.METASTASIS'

P value = 0.997 (Chi-square test), Q value = 1

Table S67. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'DISTANT.METASTASIS'

nPatients	M0	M1	MX
ALL	80	5	51
subtype1	18	1	12
subtype2	33	2	22
subtype3	29	2	17

Figure S60. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'DISTANT.METASTASIS'

'MIRSEQ CNMF' versus 'LYMPH.NODE.METASTASIS'

P value = 0.0185 (Chi-square test), Q value = 1

Table S68. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N2	N3	NX
ALL	77	14	28	5	10
subtype1	15	6	6	2	2
subtype2	36	6	5	3	6
subtype3	26	2	17	0	2

Figure S61. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

'MIRSEQ CNMF' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.303 (ANOVA), Q value = 1

Table S69. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	100	1.9 (3.7)
subtype1	24	2.5 (5.0)
subtype2	38	1.1 (2.6)
subtype3	38	2.2 (3.7)

Figure S62. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

'MIRSEQ CNMF' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.113 (Chi-square test), Q value = 1

Table S70. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	37	45	49
subtype1	1	6	8	15
subtype2	0	20	21	14
subtype3	0	11	16	20

Figure S63. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #8: 'MIRSEQ CHIERARCHICAL'

Table S71. Get Full Table Description of clustering approach #8: 'MIRSEQ CHIERARCHICAL'

Cluster Labels	1	2	3	4	5
Number of samples	20	66	7	36	8

'MIRSEQ CHIERARCHICAL' versus 'Time to Death'

P value = 0.612 (logrank test), Q value = 1

Table S72. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	130	35	0.1 - 131.2 (7.0)
subtype1	16	3	1.5 - 118.9 (5.9)
subtype2	63	21	0.1 - 49.9 (7.3)
subtype3	7	1	0.8 - 7.9 (4.1)
subtype4	36	8	0.5 - 131.2 (8.5)
subtype5	8	2	1.4 - 75.3 (6.2)

Figure S64. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CHIERARCHICAL' versus 'AGE'

P value = 0.179 (ANOVA), Q value = 1

Table S73. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	136	67.6 (10.6)
subtype1	19	68.4 (12.3)
subtype2	66	65.5 (10.5)
subtype3	7	68.0 (7.2)
subtype4	36	69.5 (10.3)
subtype5	8	73.4 (8.3)

Figure S65. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'AGE'

'MIRSEQ CHIERARCHICAL' versus 'GENDER'

P value = 0.0441 (Chi-square test), Q value = 1

Table S74. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	35	102
subtype1	3	17
subtype2	18	48
subtype3	0	7
subtype4	14	22
subtype5	0	8

Figure S66. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'GENDER'

'MIRSEQ CHIERARCHICAL' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.835 (ANOVA), Q value = 1

Table S75. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

	nPatients	Mean (Std.Dev)
ALL	39	78.2 (16.0)
subtype1	5	82.0 (11.0)
subtype2	27	77.0 (17.9)
subtype3	2	85.0 (7.1)
subtype4	4	77.5 (15.0)
subtype5	1	80.0 (NA)

Figure S67. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

'MIRSEQ CHIERARCHICAL' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.247 (ANOVA), Q value = 1

Table S76. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	85	36.8 (22.9)
subtype1	12	37.5 (25.8)
subtype2	44	36.6 (19.9)
subtype3	5	53.2 (37.0)
subtype4	17	38.1 (24.1)
subtype5	7	22.3 (18.9)

Figure S68. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

'MIRSEQ CHIERARCHICAL' versus 'DISTANT.METASTASIS'

P value = 0.914 (Chi-square test), Q value = 1

Table S77. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'DISTANT.METASTASIS'

nPatients	M0	M1	MX
ALL	80	5	51
subtype1	11	1	8
subtype2	39	3	24
subtype3	2	0	4
subtype4	23	1	12
subtype5	5	0	3

Figure S69. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'DISTANT.METASTASIS'

'MIRSEQ CHIERARCHICAL' versus 'LYMPH.NODE.METASTASIS'

P value = 0.0458 (Chi-square test), Q value = 1

Table S78. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N2	N3	NX
ALL	77	14	28	5	10
subtype1	10	3	4	1	2
subtype2	41	5	8	3	8
subtype3	2	2	1	1	0
subtype4	19	2	14	0	0
subtype5	5	2	1	0	0

Figure S70. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

'MIRSEQ CHIERARCHICAL' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.016 (ANOVA), Q value = 1

Table S79. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	100	1.9 (3.7)
subtype1	14	1.6 (3.2)
subtype2	42	1.0 (2.3)
subtype3	5	6.4 (9.3)
subtype4	31	2.7 (4.1)
subtype5	8	0.8 (1.4)

Figure S71. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

'MIRSEQ CHIERARCHICAL' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.178 (Chi-square test), Q value = 1

Table S80. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	37	45	49
subtype1	1	4	5	8
subtype2	0	24	24	16
subtype3	0	1	2	4
subtype4	0	7	11	17
subtype5	0	1	3	4

Figure S72. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #9: 'MIRseq Mature CNMF subtypes'

Table S81. Get Full Table Description of clustering approach #9: 'MIRseq Mature CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	28	47	62

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

P value = 0.929 (logrank test), Q value = 1

Table S82. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	130	35	0.1 - 131.2 (7.0)
subtype1	24	4	0.8 - 118.9 (5.3)
subtype2	47	14	0.4 - 131.2 (7.3)
subtype3	59	17	0.1 - 49.9 (7.0)

Figure S73. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature CNMF subtypes' versus 'AGE'

P value = 0.187 (ANOVA), Q value = 1

Table S83. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	136	67.6 (10.6)
subtype1	27	69.7 (11.8)
subtype2	47	68.7 (9.6)
subtype3	62	65.8 (10.6)

Figure S74. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'AGE'

'MIRseq Mature CNMF subtypes' versus 'GENDER'

P value = 0.0101 (Fisher's exact test), Q value = 0.86

Table S84. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	35	102
subtype1	3	25
subtype2	19	28
subtype3	13	49

Figure S75. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'GENDER'

'MIRseq Mature CNMF subtypes' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.336 (ANOVA), Q value = 1

Table S85. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

	nPatients	Mean (Std.Dev)
ALL	39	78.2 (16.0)
subtype1	6	83.3 (10.3)
subtype2	6	70.0 (19.0)
subtype3	27	78.9 (16.3)

Figure S76. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

'MIRseq Mature CNMF subtypes' versus 'NUMBERPACKYEARSSMOKED'

P value = 1 (ANOVA), Q value = 1

Table S86. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	85	36.8 (22.9)
subtype1	18	36.7 (29.4)
subtype2	24	36.9 (25.9)
subtype3	43	36.8 (18.2)

Figure S77. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

'MIRseq Mature CNMF subtypes' versus 'DISTANT.METASTASIS'

P value = 0.938 (Chi-square test), Q value = 1

Table S87. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

nPatients	M0	M1	MX
ALL	80	5	51
subtype1	15	1	11
subtype2	26	2	19
subtype3	39	2	21

Figure S78. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

'MIRseq Mature CNMF subtypes' versus 'LYMPH.NODE.METASTASIS'

P value = 0.00713 (Chi-square test), Q value = 0.61

Table S88. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N2	N3	NX
ALL	77	14	28	5	10
subtype1	12	5	6	2	2
subtype2	26	2	17	0	1
subtype3	39	7	5	3	7

Figure S79. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

'MIRseq Mature CNMF subtypes' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.112 (ANOVA), Q value = 1

Table S89. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	100	1.9 (3.7)
subtype1	21	2.8 (5.3)
subtype2	39	2.3 (3.7)
subtype3	40	0.9 (2.4)

Figure S80. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

'MIRseq Mature CNMF subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.05 (Chi-square test), Q value = 1

Table S90. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	37	45	49
subtype1	1	5	7	13
subtype2	0	10	14	21
subtype3	0	22	24	15

Figure S81. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #10: 'MIRseq Mature cHierClus subtypes'

Table S91. Get Full Table Description of clustering approach #10: 'MIRseq Mature cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	24	35	78

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

P value = 0.115 (logrank test), Q value = 1

Table S92. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	130	35	0.1 - 131.2 (7.0)
subtype1	21	1	0.8 - 118.9 (5.2)
subtype2	35	9	0.5 - 131.2 (8.3)
subtype3	74	25	0.1 - 49.9 (7.1)

Figure S82. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature cHierClus subtypes' versus 'AGE'

P value = 0.282 (ANOVA), Q value = 1

Table S93. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	136	67.6 (10.6)
subtype1	23	70.2 (9.7)
subtype2	35	68.4 (9.6)
subtype3	78	66.4 (11.2)

Figure S83. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'MIRseq Mature cHierClus subtypes' versus 'GENDER'

P value = 0.1 (Fisher's exact test), Q value = 1

Table S94. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	35	102
subtype1	3	21
subtype2	13	22
subtype3	19	59

Figure S84. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

'MIRseq Mature cHierClus subtypes' versus 'KARNOFSKY.PERFORMANCE.SCORE'

P value = 0.54 (ANOVA), Q value = 1

Table S95. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

	nPatients	Mean (Std.Dev)
ALL	39	78.2 (16.0)
subtype1	6	85.0 (8.4)
subtype2	4	77.5 (15.0)
subtype3	29	76.9 (17.3)

Figure S85. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'KARNOFSKY.PERFORMANCE.SCORE'

'MIRseq Mature cHierClus subtypes' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.917 (ANOVA), Q value = 1

Table S96. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	85	36.8 (22.9)
subtype1	15	38.4 (24.6)
subtype2	18	37.8 (23.0)
subtype3	52	36.0 (22.8)

Figure S86. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'NUMBERPACKYEARSSMOKED'

'MIRseq Mature cHierClus subtypes' versus 'DISTANT.METASTASIS'

P value = 0.953 (Chi-square test), Q value = 1

Table S97. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

nPatients	M0	M1	MX
ALL	80	5	51
subtype1	12	1	10
subtype2	22	1	12
subtype3	46	3	29

Figure S87. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'DISTANT.METASTASIS'

'MIRseq Mature cHierClus subtypes' versus 'LYMPH.NODE.METASTASIS'

P value = 0.00213 (Chi-square test), Q value = 0.19

Table S98. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N2	N3	NX
ALL	77	14	28	5	10
subtype1	9	6	4	2	2
subtype2	18	2	14	0	0
subtype3	50	6	10	3	8

Figure S88. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'LYMPH.NODE.METASTASIS'

'MIRseq Mature cHierClus subtypes' versus 'NUMBER.OF.LYMPH.NODES'

P value = 0.0634 (ANOVA), Q value = 1

Table S99. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

	nPatients	Mean (Std.Dev)
ALL	100	1.9 (3.7)
subtype1	17	3.1 (5.8)
subtype2	30	2.6 (4.0)
subtype3	53	1.1 (2.3)

Figure S89. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'NUMBER.OF.LYMPH.NODES'

'MIRseq Mature cHierClus subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.0221 (Chi-square test), Q value = 1

Table S100. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IV
ALL	1	37	45	49
subtype1	1	4	5	12
subtype2	0	8	9	17
subtype3	0	25	31	20

Figure S90. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'NEOPLASM.DISEASESTAGE'

Methods & Data

Input

Cluster data file = BLCA-TP.mergedcluster.txt
Clinical data file = BLCA-TP.clin.merged.picked.txt
Number of patients = 137
Number of clustering approaches = 10
Number of selected clinical features = 10
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

ANOVA analysis

For continuous numerical clinical features, one-way analysis of variance (Howell 2002) was applied to compare the clinical values between tumor subtypes using 'anova' 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

Chi-square test

For multi-class clinical features (nominal or ordinal), Chi-square tests (Greenwood and Nikulin 1996) were used to estimate the P values using the 'chisq.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

This is an experimental feature. The full results of the analysis summarized in this report can be downloaded from the TCGA Data Coordination Center.

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] Howell, D, Statistical Methods for Psychology. (5th ed.), Duxbury Press:324-5 (2002)

[5] 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)

[6] Greenwood and Nikulin, A guide to chi-squared testing, Wiley, New York. ISBN 047155779X (1996)

[7] 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