Correlation between aggregated molecular cancer subtypes and selected clinical features

Skin Cutaneous Melanoma (Metastatic)

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/C1F47M68

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 8 clinical features across 180 patients, 4 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 do not correlate to any clinical features.
3 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 5 subtypes that do not correlate to any clinical features.
CNMF clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that correlate to 'Time to Death' and 'PRIMARY.SITE.OF.DISEASE'.
Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that correlate to 'Time to Death' and 'PRIMARY.SITE.OF.DISEASE'.
3 subtypes identified in current cancer cohort by 'MIRSEQ CNMF'. These subtypes do not correlate to any clinical features.
3 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 do not correlate to any clinical features.

Results

Overview of the results

Table 1. Get Full Table Overview of the association between subtypes identified by 10 different clustering approaches and 8 clinical features. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, 4 significant findings detected.


Clinical Features	Time to Death	AGE	PRIMARY SITE OF DISEASE	GENDER	DISTANT METASTASIS	LYMPH NODE METASTASIS	TUMOR STAGECODE	NEOPLASM DISEASESTAGE
Statistical Tests	logrank test	ANOVA	Chi-square test	Fisher's exact test	Chi-square test	Chi-square test	ANOVA	Chi-square test
Copy Number Ratio CNMF subtypes	0.168 (1.00)	0.33 (1.00)	0.225 (1.00)	0.194 (1.00)	0.0985 (1.00)	0.171 (1.00)		0.269 (1.00)
METHLYATION CNMF	0.0906 (1.00)	0.132 (1.00)	0.0969 (1.00)	0.108 (1.00)	0.535 (1.00)	0.208 (1.00)		0.318 (1.00)
RPPA CNMF subtypes	0.0712 (1.00)	0.296 (1.00)	0.0208 (1.00)	0.241 (1.00)	0.462 (1.00)	0.0296 (1.00)		0.506 (1.00)
RPPA cHierClus subtypes	0.0792 (1.00)	0.869 (1.00)	0.0725 (1.00)	0.662 (1.00)	0.504 (1.00)	0.0437 (1.00)		0.0482 (1.00)
RNAseq CNMF subtypes	0.00215 (0.144)	0.0623 (1.00)	1.46e-05 (0.00101)	0.624 (1.00)	0.647 (1.00)	0.465 (1.00)		0.333 (1.00)
RNAseq cHierClus subtypes	0.000164 (0.0111)	0.0413 (1.00)	1.64e-06 (0.000115)	0.51 (1.00)	0.356 (1.00)	0.925 (1.00)		0.0599 (1.00)
MIRSEQ CNMF	0.0486 (1.00)	0.254 (1.00)	0.00608 (0.401)	0.6 (1.00)	0.685 (1.00)	0.313 (1.00)		0.251 (1.00)
MIRSEQ CHIERARCHICAL	0.598 (1.00)	0.442 (1.00)	0.896 (1.00)	0.281 (1.00)	0.295 (1.00)	0.272 (1.00)		0.226 (1.00)
MIRseq Mature CNMF subtypes	0.0395 (1.00)	0.0317 (1.00)	0.0537 (1.00)	0.186 (1.00)	0.426 (1.00)	0.255 (1.00)		0.0532 (1.00)
MIRseq Mature cHierClus subtypes	0.807 (1.00)	0.132 (1.00)	0.791 (1.00)	0.17 (1.00)	0.282 (1.00)	0.119 (1.00)		0.723 (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	57	66	57

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

P value = 0.168 (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	177	91	0.2 - 357.4 (47.5)
subtype1	57	30	0.9 - 357.4 (48.9)
subtype2	65	36	0.2 - 248.6 (40.3)
subtype3	55	25	4.2 - 314.5 (53.3)

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 'AGE'

P value = 0.33 (ANOVA), Q value = 1

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

	nPatients	Mean (Std.Dev)
ALL	179	56.0 (16.0)
subtype1	57	53.5 (16.5)
subtype2	66	56.8 (16.8)
subtype3	56	57.7 (14.4)

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 'PRIMARY.SITE.OF.DISEASE'

P value = 0.225 (Chi-square test), Q value = 1

Table S4. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	27	1	36	116
subtype1	9	1	12	35
subtype2	14	0	14	38
subtype3	4	0	10	43

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

'Copy Number Ratio CNMF subtypes' versus 'GENDER'

P value = 0.194 (Fisher's exact test), Q value = 1

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

nPatients	FEMALE	MALE
ALL	70	110
subtype1	23	34
subtype2	30	36
subtype3	17	40

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

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

P value = 0.0985 (Chi-square test), Q value = 1

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

nPatients	M0	M1	M1A	M1B	M1C
ALL	157	2	2	2	2
subtype1	47	0	1	2	2
subtype2	57	2	1	0	0
subtype3	53	0	0	0	0

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

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

P value = 0.171 (Chi-square test), Q value = 1

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

nPatients	N0	N1	N1A	N1B	N2	N2A	N2B	N2C	N3	NX
ALL	98	2	8	16	1	4	12	5	18	2
subtype1	28	0	3	3	0	2	3	3	8	2
subtype2	34	0	4	10	0	1	4	2	6	0
subtype3	36	2	1	3	1	1	5	0	4	0

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

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

P value = 0.269 (Chi-square test), Q value = 1

Table S8. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	4	18	10	15	19	9	10	8	10	6	19	25	6
subtype1	3	3	3	6	4	2	2	4	3	2	6	8	3
subtype2	0	3	4	6	9	2	4	3	2	3	9	10	3
subtype3	1	12	3	3	6	5	4	1	5	1	4	7	0

Figure S7. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #2: 'METHLYATION CNMF'

Table S9. Get Full Table Description of clustering approach #2: 'METHLYATION CNMF'

Cluster Labels	1	2	3
Number of samples	50	69	61

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.0906 (logrank test), Q value = 1

Table S10. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	177	91	0.2 - 357.4 (47.5)
subtype1	50	31	0.2 - 204.6 (40.9)
subtype2	68	31	0.9 - 357.4 (49.5)
subtype3	59	29	2.7 - 346.0 (47.3)

Figure S8. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'AGE'

P value = 0.132 (ANOVA), Q value = 1

Table S11. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	179	56.0 (16.0)
subtype1	50	55.8 (16.5)
subtype2	68	58.8 (16.6)
subtype3	61	53.2 (14.6)

Figure S9. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

'METHLYATION CNMF' versus 'PRIMARY.SITE.OF.DISEASE'

P value = 0.0969 (Chi-square test), Q value = 1

Table S12. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	27	1	36	116
subtype1	12	0	12	26
subtype2	11	0	14	44
subtype3	4	1	10	46

Figure S10. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'METHLYATION CNMF' versus 'GENDER'

P value = 0.108 (Fisher's exact test), Q value = 1

Table S13. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	70	110
subtype1	19	31
subtype2	33	36
subtype3	18	43

Figure S11. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'GENDER'

'METHLYATION CNMF' versus 'DISTANT.METASTASIS'

P value = 0.535 (Chi-square test), Q value = 1

Table S14. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'DISTANT.METASTASIS'

nPatients	M0	M1	M1A	M1B	M1C
ALL	157	2	2	2	2
subtype1	43	1	1	0	0
subtype2	61	1	0	1	0
subtype3	53	0	1	1	2

Figure S12. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'DISTANT.METASTASIS'

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

P value = 0.208 (Chi-square test), Q value = 1

Table S15. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N1A	N1B	N2	N2A	N2B	N2C	N3	NX
ALL	98	2	8	16	1	4	12	5	18	2
subtype1	29	0	4	7	0	0	3	1	2	0
subtype2	38	0	3	1	1	3	4	3	9	1
subtype3	31	2	1	8	0	1	5	1	7	1

Figure S13. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

'METHLYATION CNMF' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.318 (Chi-square test), Q value = 1

Table S16. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	4	18	10	15	19	9	10	8	10	6	19	25	6
subtype1	1	3	1	4	8	2	4	4	4	2	4	7	2
subtype2	3	4	7	5	5	4	5	4	2	3	8	9	1
subtype3	0	11	2	6	6	3	1	0	4	1	7	9	3

Figure S14. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #3: 'RPPA CNMF subtypes'

Table S17. Get Full Table Description of clustering approach #3: 'RPPA CNMF subtypes'

Cluster Labels	1	2	3	4
Number of samples	32	16	27	34

'RPPA CNMF subtypes' versus 'Time to Death'

P value = 0.0712 (logrank test), Q value = 1

Table S18. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	108	58	2.6 - 357.4 (47.5)
subtype1	32	23	10.1 - 204.6 (50.5)
subtype2	16	3	4.2 - 346.0 (46.6)
subtype3	26	14	2.7 - 248.6 (34.1)
subtype4	34	18	2.6 - 357.4 (50.3)

Figure S15. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA CNMF subtypes' versus 'AGE'

P value = 0.296 (ANOVA), Q value = 1

Table S19. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	109	56.4 (16.5)
subtype1	32	54.4 (16.1)
subtype2	16	61.8 (15.7)
subtype3	27	59.0 (17.6)
subtype4	34	53.7 (16.2)

Figure S16. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'AGE'

'RPPA CNMF subtypes' versus 'PRIMARY.SITE.OF.DISEASE'

P value = 0.0208 (Chi-square test), Q value = 1

Table S20. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	13	1	15	80
subtype1	4	0	9	19
subtype2	0	1	0	15
subtype3	5	0	4	18
subtype4	4	0	2	28

Figure S17. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'RPPA CNMF subtypes' versus 'GENDER'

P value = 0.241 (Fisher's exact test), Q value = 1

Table S21. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	44	65
subtype1	13	19
subtype2	10	6
subtype3	10	17
subtype4	11	23

Figure S18. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'GENDER'

'RPPA CNMF subtypes' versus 'DISTANT.METASTASIS'

P value = 0.462 (Chi-square test), Q value = 1

Table S22. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

nPatients	M0	M1	M1B	M1C
ALL	90	1	2	2
subtype1	27	0	1	2
subtype2	15	0	0	0
subtype3	24	0	1	0
subtype4	24	1	0	0

Figure S19. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

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

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

Table S23. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N1A	N1B	N2A	N2B	N2C	N3	NX
ALL	56	2	3	13	2	5	4	9	2
subtype1	14	1	3	5	0	1	2	4	0
subtype2	11	1	0	0	0	2	0	0	1
subtype3	9	0	0	6	2	2	2	3	1
subtype4	22	0	0	2	0	0	0	2	0

Figure S20. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

'RPPA CNMF subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.506 (Chi-square test), Q value = 1

Table S24. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	3	11	8	8	10	5	5	3	6	2	11	13	4
subtype1	1	2	1	3	2	2	2	0	3	2	4	5	3
subtype2	0	2	2	2	1	1	1	1	1	0	1	1	0
subtype3	1	2	2	0	2	0	2	0	1	0	5	6	0
subtype4	1	5	3	3	5	2	0	2	1	0	1	1	1

Figure S21. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #4: 'RPPA cHierClus subtypes'

Table S25. Get Full Table Description of clustering approach #4: 'RPPA cHierClus subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	20	15	35	22	17

'RPPA cHierClus subtypes' versus 'Time to Death'

P value = 0.0792 (logrank test), Q value = 1

Table S26. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	108	58	2.6 - 357.4 (47.5)
subtype1	20	14	6.4 - 357.4 (50.5)
subtype2	15	3	4.2 - 346.0 (54.7)
subtype3	35	19	2.6 - 216.9 (53.2)
subtype4	21	14	13.9 - 117.2 (39.6)
subtype5	17	8	2.7 - 248.6 (36.0)

Figure S22. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA cHierClus subtypes' versus 'AGE'

P value = 0.869 (ANOVA), Q value = 1

Table S27. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	109	56.4 (16.5)
subtype1	20	53.8 (15.7)
subtype2	15	57.9 (13.3)
subtype3	35	55.5 (16.7)
subtype4	22	56.9 (17.9)
subtype5	17	59.4 (18.9)

Figure S23. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'RPPA cHierClus subtypes' versus 'PRIMARY.SITE.OF.DISEASE'

P value = 0.0725 (Chi-square test), Q value = 1

Table S28. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	13	1	15	80
subtype1	2	0	4	14
subtype2	1	0	0	14
subtype3	5	0	6	24
subtype4	5	0	5	12
subtype5	0	1	0	16

Figure S24. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'RPPA cHierClus subtypes' versus 'GENDER'

P value = 0.662 (Chi-square test), Q value = 1

Table S29. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	44	65
subtype1	6	14
subtype2	8	7
subtype3	14	21
subtype4	8	14
subtype5	8	9

Figure S25. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

'RPPA cHierClus subtypes' versus 'DISTANT.METASTASIS'

P value = 0.504 (Chi-square test), Q value = 1

Table S30. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

nPatients	M0	M1	M1B	M1C
ALL	90	1	2	2
subtype1	13	1	1	0
subtype2	12	0	0	0
subtype3	29	0	1	0
subtype4	20	0	0	1
subtype5	16	0	0	1

Figure S26. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

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

P value = 0.0437 (Chi-square test), Q value = 1

Table S31. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N1A	N1B	N2A	N2B	N2C	N3	NX
ALL	56	2	3	13	2	5	4	9	2
subtype1	13	0	0	0	0	0	0	3	0
subtype2	9	0	0	1	0	1	0	0	1
subtype3	17	1	0	5	0	1	2	4	0
subtype4	8	0	3	6	0	1	1	2	0
subtype5	9	1	0	1	2	2	1	0	1

Figure S27. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

'RPPA cHierClus subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.0482 (Chi-square test), Q value = 1

Table S32. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	3	11	8	8	10	5	5	3	6	2	11	13	4
subtype1	2	2	0	1	1	3	1	2	0	0	0	1	2
subtype2	0	4	2	1	0	1	0	1	0	0	2	0	0
subtype3	1	1	4	3	6	0	2	0	4	0	2	5	0
subtype4	0	2	1	2	0	1	1	0	1	2	5	5	1
subtype5	0	2	1	1	3	0	1	0	1	0	2	2	1

Figure S28. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #5: 'RNAseq CNMF subtypes'

Table S33. Get Full Table Description of clustering approach #5: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	57	46	68

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.00215 (logrank test), Q value = 0.14

Table S34. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	168	87	0.2 - 357.4 (47.5)
subtype1	57	35	7.8 - 357.4 (61.2)
subtype2	46	14	4.2 - 203.0 (54.9)
subtype3	65	38	0.2 - 228.6 (36.0)

Figure S29. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'AGE'

P value = 0.0623 (ANOVA), Q value = 1

Table S35. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	170	56.1 (16.2)
subtype1	57	52.3 (15.5)
subtype2	46	56.1 (17.4)
subtype3	67	59.2 (15.5)

Figure S30. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq CNMF subtypes' versus 'PRIMARY.SITE.OF.DISEASE'

P value = 1.46e-05 (Chi-square test), Q value = 0.001

Table S36. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	25	1	33	112
subtype1	7	0	22	28
subtype2	3	0	1	42
subtype3	15	1	10	42

Figure S31. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 0.624 (Fisher's exact test), Q value = 1

Table S37. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	66	105
subtype1	24	33
subtype2	15	31
subtype3	27	41

Figure S32. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'GENDER'

'RNAseq CNMF subtypes' versus 'DISTANT.METASTASIS'

P value = 0.647 (Chi-square test), Q value = 1

Table S38. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

nPatients	M0	M1	M1A	M1B	M1C
ALL	148	2	2	2	2
subtype1	48	0	1	1	1
subtype2	42	0	0	0	1
subtype3	58	2	1	1	0

Figure S33. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

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

P value = 0.465 (Chi-square test), Q value = 1

Table S39. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N1A	N1B	N2	N2A	N2B	N2C	N3	NX
ALL	95	2	6	16	1	4	10	5	16	2
subtype1	29	1	3	4	0	1	3	4	5	1
subtype2	31	1	1	2	0	0	3	0	4	1
subtype3	35	0	2	10	1	3	4	1	7	0

Figure S34. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

'RNAseq CNMF subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.333 (Chi-square test), Q value = 1

Table S40. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	4	17	10	14	19	8	10	8	9	5	17	23	6
subtype1	2	8	1	4	7	2	3	1	5	2	5	7	3
subtype2	0	6	5	4	8	4	1	1	2	0	4	5	1
subtype3	2	3	4	6	4	2	6	6	2	3	8	11	2

Figure S35. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #6: 'RNAseq cHierClus subtypes'

Table S41. Get Full Table Description of clustering approach #6: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	46	42	83

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.000164 (logrank test), Q value = 0.011

Table S42. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	168	87	0.2 - 357.4 (47.5)
subtype1	45	30	0.2 - 228.6 (35.9)
subtype2	42	28	7.8 - 357.4 (56.8)
subtype3	81	29	2.7 - 346.0 (49.5)

Figure S36. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'AGE'

P value = 0.0413 (ANOVA), Q value = 1

Table S43. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	170	56.1 (16.2)
subtype1	45	60.9 (16.1)
subtype2	42	52.4 (15.3)
subtype3	83	55.3 (16.3)

Figure S37. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq cHierClus subtypes' versus 'PRIMARY.SITE.OF.DISEASE'

P value = 1.64e-06 (Chi-square test), Q value = 0.00012

Table S44. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	25	1	33	112
subtype1	14	0	10	22
subtype2	5	0	17	20
subtype3	6	1	6	70

Figure S38. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 0.51 (Fisher's exact test), Q value = 1

Table S45. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	66	105
subtype1	21	25
subtype2	16	26
subtype3	29	54

Figure S39. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

'RNAseq cHierClus subtypes' versus 'DISTANT.METASTASIS'

P value = 0.356 (Chi-square test), Q value = 1

Table S46. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

nPatients	M0	M1	M1A	M1B	M1C
ALL	148	2	2	2	2
subtype1	38	2	1	0	0
subtype2	34	0	0	1	1
subtype3	76	0	1	1	1

Figure S40. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

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

P value = 0.925 (Chi-square test), Q value = 1

Table S47. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N1A	N1B	N2	N2A	N2B	N2C	N3	NX
ALL	95	2	6	16	1	4	10	5	16	2
subtype1	26	1	1	5	0	0	3	1	5	0
subtype2	22	0	3	3	0	1	1	2	4	0
subtype3	47	1	2	8	1	3	6	2	7	2

Figure S41. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

'RNAseq cHierClus subtypes' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.0599 (Chi-square test), Q value = 1

Table S48. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	4	17	10	14	19	8	10	8	9	5	17	23	6
subtype1	1	2	1	1	5	2	6	6	1	1	4	9	3
subtype2	2	5	0	5	5	2	2	1	2	2	3	5	2
subtype3	1	10	9	8	9	4	2	1	6	2	10	9	1

Figure S42. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #7: 'MIRSEQ CNMF'

Table S49. Get Full Table Description of clustering approach #7: 'MIRSEQ CNMF'

Cluster Labels	1	2	3
Number of samples	44	74	51

'MIRSEQ CNMF' versus 'Time to Death'

P value = 0.0486 (logrank test), Q value = 1

Table S50. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	166	86	0.2 - 357.4 (47.5)
subtype1	43	16	0.2 - 357.4 (46.8)
subtype2	74	42	2.6 - 216.9 (42.4)
subtype3	49	28	7.8 - 314.5 (61.0)

Figure S43. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CNMF' versus 'AGE'

P value = 0.254 (ANOVA), Q value = 1

Table S51. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	168	56.1 (16.0)
subtype1	44	57.2 (15.1)
subtype2	74	57.6 (16.3)
subtype3	50	52.9 (16.2)

Figure S44. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'AGE'

'MIRSEQ CNMF' versus 'PRIMARY.SITE.OF.DISEASE'

P value = 0.00608 (Chi-square test), Q value = 0.4

Table S52. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	24	1	34	110
subtype1	4	0	7	33
subtype2	11	1	8	54
subtype3	9	0	19	23

Figure S45. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'MIRSEQ CNMF' versus 'GENDER'

P value = 0.6 (Fisher's exact test), Q value = 1

Table S53. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	64	105
subtype1	19	25
subtype2	28	46
subtype3	17	34

Figure S46. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'GENDER'

'MIRSEQ CNMF' versus 'DISTANT.METASTASIS'

P value = 0.685 (Chi-square test), Q value = 1

Table S54. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'DISTANT.METASTASIS'

nPatients	M0	M1	M1A	M1B	M1C
ALL	146	2	2	2	2
subtype1	32	1	1	1	1
subtype2	66	1	1	1	0
subtype3	48	0	0	0	1

Figure S47. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'DISTANT.METASTASIS'

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

P value = 0.313 (Chi-square test), Q value = 1

Table S55. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N1A	N1B	N2	N2A	N2B	N2C	N3	NX
ALL	92	2	7	16	1	4	11	5	15	2
subtype1	19	0	1	4	0	2	2	1	6	2
subtype2	42	1	2	8	0	2	6	1	7	0
subtype3	31	1	4	4	1	0	3	3	2	0

Figure S48. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

'MIRSEQ CNMF' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.251 (Chi-square test), Q value = 1

Table S56. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	4	17	10	14	18	7	10	7	9	6	18	22	6
subtype1	2	4	3	1	4	1	0	2	1	1	6	6	3
subtype2	1	5	6	5	11	3	7	2	2	2	8	11	2
subtype3	1	8	1	8	3	3	3	3	6	3	4	5	1

Figure S49. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #8: 'MIRSEQ CHIERARCHICAL'

Table S57. Get Full Table Description of clustering approach #8: 'MIRSEQ CHIERARCHICAL'

Cluster Labels	1	2	3
Number of samples	21	105	43

'MIRSEQ CHIERARCHICAL' versus 'Time to Death'

P value = 0.598 (logrank test), Q value = 1

Table S58. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	166	86	0.2 - 357.4 (47.5)
subtype1	20	7	0.2 - 357.4 (28.7)
subtype2	103	54	2.6 - 314.5 (47.3)
subtype3	43	25	7.8 - 346.0 (60.2)

Figure S50. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CHIERARCHICAL' versus 'AGE'

P value = 0.442 (ANOVA), Q value = 1

Table S59. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	168	56.1 (16.0)
subtype1	21	57.7 (16.5)
subtype2	104	56.8 (16.7)
subtype3	43	53.4 (14.0)

Figure S51. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'AGE'

'MIRSEQ CHIERARCHICAL' versus 'PRIMARY.SITE.OF.DISEASE'

P value = 0.896 (Chi-square test), Q value = 1

Table S60. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	24	1	34	110
subtype1	3	0	5	13
subtype2	16	1	18	70
subtype3	5	0	11	27

Figure S52. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'MIRSEQ CHIERARCHICAL' versus 'GENDER'

P value = 0.281 (Fisher's exact test), Q value = 1

Table S61. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	64	105
subtype1	8	13
subtype2	44	61
subtype3	12	31

Figure S53. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'GENDER'

'MIRSEQ CHIERARCHICAL' versus 'DISTANT.METASTASIS'

P value = 0.295 (Chi-square test), Q value = 1

Table S62. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'DISTANT.METASTASIS'

nPatients	M0	M1	M1A	M1B	M1C
ALL	146	2	2	2	2
subtype1	16	1	1	1	0
subtype2	91	1	1	1	1
subtype3	39	0	0	0	1

Figure S54. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'DISTANT.METASTASIS'

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

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

Table S63. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N1A	N1B	N2	N2A	N2B	N2C	N3	NX
ALL	92	2	7	16	1	4	11	5	15	2
subtype1	8	0	1	3	0	2	2	0	2	1
subtype2	62	1	4	9	0	2	8	2	8	0
subtype3	22	1	2	4	1	0	1	3	5	1

Figure S55. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

'MIRSEQ CHIERARCHICAL' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.226 (Chi-square test), Q value = 1

Table S64. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	4	17	10	14	18	7	10	7	9	6	18	22	6
subtype1	1	0	1	0	3	1	0	1	0	1	6	2	2
subtype2	2	9	8	10	12	5	8	4	4	3	10	13	3
subtype3	1	8	1	4	3	1	2	2	5	2	2	7	1

Figure S56. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #9: 'MIRseq Mature CNMF subtypes'

Table S65. Get Full Table Description of clustering approach #9: 'MIRseq Mature CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	52	69	48

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

P value = 0.0395 (logrank test), Q value = 1

Table S66. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	166	86	0.2 - 357.4 (47.5)
subtype1	51	20	0.2 - 357.4 (47.5)
subtype2	69	40	2.6 - 248.6 (39.6)
subtype3	46	26	6.4 - 314.5 (61.1)

Figure S57. 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 = 0.0317 (ANOVA), Q value = 1

Table S67. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	168	56.1 (16.0)
subtype1	52	58.8 (14.2)
subtype2	69	57.5 (17.2)
subtype3	47	51.0 (15.2)

Figure S58. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'AGE'

'MIRseq Mature CNMF subtypes' versus 'PRIMARY.SITE.OF.DISEASE'

P value = 0.0537 (Chi-square test), Q value = 1

Table S68. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	24	1	34	110
subtype1	8	0	9	35
subtype2	9	1	8	51
subtype3	7	0	17	24

Figure S59. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'MIRseq Mature CNMF subtypes' versus 'GENDER'

P value = 0.186 (Fisher's exact test), Q value = 1

Table S69. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	64	105
subtype1	25	27
subtype2	24	45
subtype3	15	33

Figure S60. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'GENDER'

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

P value = 0.426 (Chi-square test), Q value = 1

Table S70. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

nPatients	M0	M1	M1A	M1B	M1C
ALL	146	2	2	2	2
subtype1	41	1	1	1	0
subtype2	62	0	1	1	0
subtype3	43	1	0	0	2

Figure S61. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

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

P value = 0.255 (Chi-square test), Q value = 1

Table S71. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N1A	N1B	N2	N2A	N2B	N2C	N3	NX
ALL	92	2	7	16	1	4	11	5	15	2
subtype1	25	0	2	5	0	2	1	1	7	2
subtype2	39	1	1	8	0	2	6	1	6	0
subtype3	28	1	4	3	1	0	4	3	2	0

Figure S62. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

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

P value = 0.0532 (Chi-square test), Q value = 1

Table S72. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	4	17	10	14	18	7	10	7	9	6	18	22	6
subtype1	2	5	5	1	5	1	1	4	1	1	7	8	2
subtype2	1	4	4	5	11	3	7	1	2	2	7	11	1
subtype3	1	8	1	8	2	3	2	2	6	3	4	3	3

Figure S63. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Clustering Approach #10: 'MIRseq Mature cHierClus subtypes'

Table S73. Get Full Table Description of clustering approach #10: 'MIRseq Mature cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	103	51	15

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

P value = 0.807 (logrank test), Q value = 1

Table S74. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	166	86	0.2 - 357.4 (47.5)
subtype1	101	51	2.6 - 314.5 (44.4)
subtype2	51	30	7.8 - 357.4 (61.2)
subtype3	14	5	0.2 - 203.0 (26.6)

Figure S64. 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 = 0.132 (ANOVA), Q value = 1

Table S75. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	168	56.1 (16.0)
subtype1	102	57.0 (16.8)
subtype2	51	52.7 (13.6)
subtype3	15	61.0 (17.3)

Figure S65. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'MIRseq Mature cHierClus subtypes' versus 'PRIMARY.SITE.OF.DISEASE'

P value = 0.791 (Chi-square test), Q value = 1

Table S76. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	DISTANT METASTASIS	PRIMARY TUMOR	REGIONAL CUTANEOUS OR SUBCUTANEOUS TISSUE (INCLUDES SATELLITE AND IN-TRANSIT METASTASIS)	REGIONAL LYMPH NODE
ALL	24	1	34	110
subtype1	15	1	17	70
subtype2	7	0	14	30
subtype3	2	0	3	10

Figure S66. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'MIRseq Mature cHierClus subtypes' versus 'GENDER'

P value = 0.17 (Fisher's exact test), Q value = 1

Table S77. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	64	105
subtype1	43	60
subtype2	14	37
subtype3	7	8

Figure S67. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

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

P value = 0.282 (Chi-square test), Q value = 1

Table S78. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

nPatients	M0	M1	M1A	M1B	M1C
ALL	146	2	2	2	2
subtype1	90	1	1	1	1
subtype2	44	0	0	1	1
subtype3	12	1	1	0	0

Figure S68. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'DISTANT.METASTASIS'

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

P value = 0.119 (Chi-square test), Q value = 1

Table S79. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

nPatients	N0	N1	N1A	N1B	N2	N2A	N2B	N2C	N3	NX
ALL	92	2	7	16	1	4	11	5	15	2
subtype1	60	1	4	9	0	2	9	2	8	0
subtype2	26	1	2	4	1	0	2	3	6	1
subtype3	6	0	1	3	0	2	0	0	1	1

Figure S69. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'LYMPH.NODE.METASTASIS'

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

P value = 0.723 (Chi-square test), Q value = 1

Table S80. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

nPatients	I OR II NOS	STAGE I	STAGE IA	STAGE IB	STAGE II	STAGE IIA	STAGE IIB	STAGE IIC	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV
ALL	4	17	10	14	18	7	10	7	9	6	18	22	6
subtype1	2	9	8	9	12	4	8	4	4	3	11	13	3
subtype2	2	8	1	5	4	2	2	2	5	2	3	7	2
subtype3	0	0	1	0	2	1	0	1	0	1	4	2	1

Figure S70. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'NEOPLASM.DISEASESTAGE'

Methods & Data

Input

Cluster data file = SKCM-TM.mergedcluster.txt
Clinical data file = SKCM-TM.clin.merged.picked.txt
Number of patients = 180
Number of clustering approaches = 10
Number of selected clinical features = 8
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

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

Fisher's exact test

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

Q value calculation

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

Download Results

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

[6] 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)

[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