Skin Cutaneous Melanoma: Correlation between molecular cancer subtypes and selected clinical features

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

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 8 different clustering approaches and 3 clinical features across 126 patients, 2 significant findings detected with P value < 0.05.

3 subtypes identified in current cancer cohort by 'CN CNMF'. 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 5 subtypes that correlate to 'AGE'.
Consensus hierarchical clustering analysis on RPPA data identified 4 subtypes that correlate to 'Time to Death'.
CNMF clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that do not correlate to any clinical features.
Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that do not correlate to any clinical features.
CNMF clustering analysis on sequencing-based miR expression data identified 4 subtypes that do not correlate to any clinical features.
Consensus hierarchical clustering analysis on sequencing-based miR expression data identified 3 subtypes that 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 8 different clustering approaches and 3 clinical features. Shown in the table are P values from statistical tests. Thresholded by P value < 0.05, 2 significant findings detected.


Clinical Features	Time to Death	AGE	GENDER
Statistical Tests	logrank test	ANOVA	Fisher's exact test
CN CNMF	0.192	0.818	0.973
METHLYATION CNMF	0.893	0.182	0.238
RPPA CNMF subtypes	0.738	0.0494	0.677
RPPA cHierClus subtypes	0.0264	0.754	0.362
RNAseq CNMF subtypes	0.3	0.0859	0.534
RNAseq cHierClus subtypes	0.886	0.0758	0.794
MIRseq CNMF subtypes	0.193	0.526	0.841
MIRseq cHierClus subtypes	0.237	0.443	0.924

Clustering Approach #1: 'CN CNMF'

Table S1. Get Full Table Description of clustering approach #1: 'CN CNMF'

Cluster Labels	1	2	3
Number of samples	46	43	37

'CN CNMF' versus 'Time to Death'

P value = 0.192 (logrank test)

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

	nPatients	nDeath	Duration Range (Median), Month
ALL	15	8	0.2 - 131.1 (62.8)
subtype1	8	5	0.2 - 131.1 (41.8)
subtype2	5	1	10.1 - 120.5 (80.3)
subtype3	2	2	32.5 - 117.9 (75.2)

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

'CN CNMF' versus 'AGE'

P value = 0.818 (ANOVA)

Table S3. Clustering Approach #1: 'CN CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	16	56.8 (14.0)
subtype1	8	57.8 (19.2)
subtype2	5	53.4 (7.8)
subtype3	3	59.7 (3.5)

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

'CN CNMF' versus 'GENDER'

P value = 0.973 (Fisher's exact test)

Table S4. Clustering Approach #1: 'CN CNMF' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	46	80
subtype1	17	29
subtype2	15	28
subtype3	14	23

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

Clustering Approach #2: 'METHLYATION CNMF'

Table S5. Get Full Table Description of clustering approach #2: 'METHLYATION CNMF'

Cluster Labels	1	2	3
Number of samples	32	47	47

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.893 (logrank test)

Table S6. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	15	8	0.2 - 131.1 (62.8)
subtype1	6	3	0.2 - 131.1 (52.0)
subtype2	4	3	26.4 - 120.5 (90.3)
subtype3	5	2	10.1 - 84.7 (44.0)

Figure S4. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'AGE'

P value = 0.182 (ANOVA)

Table S7. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	16	56.8 (14.0)
subtype1	6	52.3 (14.9)
subtype2	4	68.0 (17.6)
subtype3	6	53.7 (6.8)

Figure S5. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

'METHLYATION CNMF' versus 'GENDER'

P value = 0.238 (Fisher's exact test)

Table S8. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	46	80
subtype1	12	20
subtype2	21	26
subtype3	13	34

Figure S6. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'GENDER'

Clustering Approach #3: 'RPPA CNMF subtypes'

Table S9. Get Full Table Description of clustering approach #3: 'RPPA CNMF subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	15	17	17	21	19

'RPPA CNMF subtypes' versus 'Time to Death'

P value = 0.738 (logrank test)

Table S10. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	13	8	10.1 - 131.1 (62.8)
subtype1	4	3	10.1 - 131.1 (48.7)
subtype2	2	2	32.5 - 64.4 (48.4)
subtype3	1	0	44.0 - 44.0 (44.0)
subtype4	6	3	26.4 - 120.5 (71.4)

Figure S7. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA CNMF subtypes' versus 'AGE'

P value = 0.0494 (ANOVA)

Table S11. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	14	56.1 (14.9)
subtype1	4	44.8 (7.4)
subtype2	3	52.3 (12.9)
subtype3	1	47.0 (NA)
subtype4	6	67.0 (14.4)

Figure S8. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'AGE'

'RPPA CNMF subtypes' versus 'GENDER'

P value = 0.677 (Chi-square test)

Table S12. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	35	54
subtype1	5	10
subtype2	5	12
subtype3	9	8
subtype4	8	13
subtype5	8	11

Figure S9. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'GENDER'

Clustering Approach #4: 'RPPA cHierClus subtypes'

Table S13. Get Full Table Description of clustering approach #4: 'RPPA cHierClus subtypes'

Cluster Labels	1	2	3	4
Number of samples	23	15	34	17

'RPPA cHierClus subtypes' versus 'Time to Death'

P value = 0.0264 (logrank test)

Table S14. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	13	8	10.1 - 131.1 (62.8)
subtype1	1	0	44.0 - 44.0 (44.0)
subtype2	4	3	10.1 - 39.6 (19.5)
subtype3	5	3	62.8 - 131.1 (117.9)
subtype4	3	2	32.5 - 84.7 (64.4)

Figure S10. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA cHierClus subtypes' versus 'AGE'

P value = 0.754 (ANOVA)

Table S15. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	14	56.1 (14.9)
subtype1	1	47.0 (NA)
subtype2	4	61.2 (18.6)
subtype3	5	56.6 (17.8)
subtype4	4	52.5 (10.5)

Figure S11. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'RPPA cHierClus subtypes' versus 'GENDER'

P value = 0.362 (Fisher's exact test)

Table S16. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	35	54
subtype1	12	11
subtype2	4	11
subtype3	14	20
subtype4	5	12

Figure S12. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

Clustering Approach #5: 'RNAseq CNMF subtypes'

Table S17. Get Full Table Description of clustering approach #5: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	43	37	44

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.3 (logrank test)

Table S18. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	14	8	0.2 - 131.1 (53.4)
subtype1	5	4	10.1 - 131.1 (32.5)
subtype2	3	0	44.0 - 84.7 (80.0)
subtype3	6	4	0.2 - 117.9 (51.2)

Figure S13. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'AGE'

P value = 0.0859 (ANOVA)

Table S19. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	15	56.3 (14.4)
subtype1	5	47.2 (10.1)
subtype2	4	53.5 (4.8)
subtype3	6	65.7 (17.0)

Figure S14. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 0.534 (Fisher's exact test)

Table S20. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	45	79
subtype1	18	25
subtype2	11	26
subtype3	16	28

Figure S15. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

Clustering Approach #6: 'RNAseq cHierClus subtypes'

Table S21. Get Full Table Description of clustering approach #6: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	31	25	68

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.886 (logrank test)

Table S22. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	14	8	0.2 - 131.1 (53.4)
subtype1	5	4	10.1 - 131.1 (32.5)
subtype2	4	2	0.2 - 117.9 (51.2)
subtype3	5	2	26.4 - 84.7 (64.4)

Figure S16. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'AGE'

P value = 0.0758 (ANOVA)

Table S23. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	15	56.3 (14.4)
subtype1	5	47.2 (10.1)
subtype2	4	68.5 (11.3)
subtype3	6	55.7 (14.8)

Figure S17. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 0.794 (Fisher's exact test)

Table S24. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	45	79
subtype1	12	19
subtype2	10	15
subtype3	23	45

Figure S18. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

Clustering Approach #7: 'MIRseq CNMF subtypes'

Table S25. Get Full Table Description of clustering approach #7: 'MIRseq CNMF subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	27	43	29	22	2

'MIRseq CNMF subtypes' versus 'Time to Death'

P value = 0.193 (logrank test)

Table S26. Clustering Approach #7: 'MIRseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	13	8	0.2 - 131.1 (62.8)
subtype1	3	1	44.0 - 131.1 (80.0)
subtype2	3	3	62.8 - 117.9 (64.4)
subtype3	4	3	10.1 - 80.3 (22.5)
subtype4	3	1	0.2 - 84.7 (26.4)

Figure S19. Get High-res Image Clustering Approach #7: 'MIRseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq CNMF subtypes' versus 'AGE'

P value = 0.526 (ANOVA)

Table S27. Clustering Approach #7: 'MIRseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	13	56.4 (14.8)
subtype1	3	46.7 (11.5)
subtype2	3	60.3 (22.5)
subtype3	4	54.5 (10.4)
subtype4	3	64.7 (15.3)

Figure S20. Get High-res Image Clustering Approach #7: 'MIRseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

'MIRseq CNMF subtypes' versus 'GENDER'

P value = 0.841 (Fisher's exact test)

Table S28. Clustering Approach #7: 'MIRseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	42	79
subtype1	10	17
subtype2	16	27
subtype3	8	21
subtype4	8	14

Figure S21. Get High-res Image Clustering Approach #7: 'MIRseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

Clustering Approach #8: 'MIRseq cHierClus subtypes'

Table S29. Get Full Table Description of clustering approach #8: 'MIRseq cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	16	79	28

'MIRseq cHierClus subtypes' versus 'Time to Death'

P value = 0.237 (logrank test)

Table S30. Clustering Approach #8: 'MIRseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	14	8	0.2 - 131.1 (63.6)
subtype1	2	0	0.2 - 84.7 (42.5)
subtype2	7	5	26.4 - 131.1 (80.3)
subtype3	5	3	10.1 - 80.0 (32.5)

Figure S22. Get High-res Image Clustering Approach #8: 'MIRseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq cHierClus subtypes' versus 'AGE'

P value = 0.443 (ANOVA)

Table S31. Clustering Approach #8: 'MIRseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	14	55.7 (14.4)
subtype1	2	56.0 (4.2)
subtype2	7	58.4 (19.5)
subtype3	5	51.8 (8.2)

Figure S23. Get High-res Image Clustering Approach #8: 'MIRseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'MIRseq cHierClus subtypes' versus 'GENDER'

P value = 0.924 (Fisher's exact test)

Table S32. Clustering Approach #8: 'MIRseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	44	79
subtype1	6	10
subtype2	29	50
subtype3	9	19

Figure S24. Get High-res Image Clustering Approach #8: 'MIRseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

Methods & Data

Input

Cluster data file = SKCM-TM.mergedcluster.txt
Clinical data file = SKCM.clin.merged.picked.txt
Number of patients = 126
Number of clustering approaches = 8
Number of selected clinical features = 3
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

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)

Made with Nozzle