Correlation between aggregated molecular cancer subtypes and selected clinical features

Overview

Introduction

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

Summary

Testing the association between subtypes identified by 8 different clustering approaches and 2 clinical features across 33 patients, no significant finding 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.
5 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. These subtypes do not correlate to any clinical features.
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 4 subtypes that do not correlate to any clinical features.
3 subtypes identified in current cancer cohort by 'MIRSEQ CNMF'. These subtypes do not correlate to any clinical features.
2 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.
2 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 8 different clustering approaches and 2 clinical features. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, no significant finding detected.


Clinical Features	Time to Death	AGE
Statistical Tests	logrank test	t-test
Copy Number Ratio CNMF subtypes	0.938 (1.00)	0.526 (1.00)
METHLYATION CNMF	0.637 (1.00)	0.0435 (0.696)
RNAseq CNMF subtypes	0.419 (1.00)	0.803 (1.00)
RNAseq cHierClus subtypes	0.263 (1.00)	0.193 (1.00)
MIRSEQ CNMF	0.691 (1.00)	0.832 (1.00)
MIRSEQ CHIERARCHICAL	0.84 (1.00)	0.151 (1.00)
MIRseq Mature CNMF subtypes	0.69 (1.00)	0.551 (1.00)
MIRseq Mature cHierClus subtypes	0.96 (1.00)	0.212 (1.00)

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

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

Cluster Labels	1	2	3
Number of samples	13	12	7

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

P value = 0.938 (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	32	19	3.7 - 85.3 (15.8)
subtype1	13	9	6.6 - 74.3 (23.5)
subtype2	12	7	4.6 - 59.7 (14.4)
subtype3	7	3	3.7 - 85.3 (5.5)

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 = 0.526 (ANOVA), Q value = 1

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

	nPatients	Mean (Std.Dev)
ALL	32	69.8 (8.5)
subtype1	13	68.2 (8.3)
subtype2	12	72.0 (8.4)
subtype3	7	68.9 (9.3)

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

Clustering Approach #2: 'METHLYATION CNMF'

Table S4. Description of clustering approach #2: 'METHLYATION CNMF'

Cluster Labels	1	2	3	4	5
Number of samples	7	8	6	4	8

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.637 (logrank test), Q value = 1

Table S5. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	33	20	3.7 - 102.4 (16.3)
subtype1	7	4	4.8 - 102.4 (27.4)
subtype2	8	6	4.6 - 69.6 (18.6)
subtype3	6	3	7.2 - 47.3 (20.9)
subtype4	4	1	5.4 - 24.0 (10.5)
subtype5	8	6	3.7 - 85.3 (15.0)

Figure S3. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'AGE'

P value = 0.0435 (ANOVA), Q value = 0.7

Table S6. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	33	69.6 (8.4)
subtype1	7	64.9 (5.6)
subtype2	8	65.2 (6.8)
subtype3	6	74.8 (8.9)
subtype4	4	69.8 (4.1)
subtype5	8	74.2 (9.5)

Figure S4. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

Clustering Approach #3: 'RNAseq CNMF subtypes'

Table S7. Description of clustering approach #3: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	11	8	13

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.419 (logrank test), Q value = 1

Table S8. Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	32	19	3.7 - 102.4 (17.2)
subtype1	11	7	4.6 - 102.4 (13.9)
subtype2	8	2	5.4 - 47.3 (20.2)
subtype3	13	10	3.7 - 85.3 (18.1)

Figure S5. Get High-res Image Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'AGE'

P value = 0.803 (ANOVA), Q value = 1

Table S9. Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	32	69.9 (8.4)
subtype1	11	69.2 (8.2)
subtype2	8	71.6 (5.4)
subtype3	13	69.4 (10.3)

Figure S6. Get High-res Image Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

Clustering Approach #4: 'RNAseq cHierClus subtypes'

Table S10. Description of clustering approach #4: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	2	3	10	8	9

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.263 (logrank test), Q value = 1

Table S11. Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	30	18	3.7 - 102.4 (17.2)
subtype2	3	1	24.0 - 59.7 (50.2)
subtype3	10	7	4.6 - 102.4 (18.2)
subtype4	8	7	3.7 - 85.3 (16.4)
subtype5	9	3	5.4 - 47.3 (15.2)

Figure S7. Get High-res Image Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'AGE'

P value = 0.193 (ANOVA), Q value = 1

Table S12. Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	30	69.7 (8.6)
subtype2	3	62.3 (8.5)
subtype3	10	67.3 (5.7)
subtype4	8	72.1 (11.1)
subtype5	9	72.8 (7.9)

Figure S8. Get High-res Image Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

Clustering Approach #5: 'MIRSEQ CNMF'

Table S13. Description of clustering approach #5: 'MIRSEQ CNMF'

Cluster Labels	1	2	3	4
Number of samples	9	14	2	8

'MIRSEQ CNMF' versus 'Time to Death'

P value = 0.691 (logrank test), Q value = 1

Table S14. Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	31	19	3.7 - 102.4 (15.2)
subtype1	9	6	4.6 - 102.4 (7.2)
subtype2	14	9	4.9 - 85.3 (23.7)
subtype4	8	4	3.7 - 31.2 (14.6)

Figure S9. Get High-res Image Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CNMF' versus 'AGE'

P value = 0.832 (ANOVA), Q value = 1

Table S15. Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	31	69.8 (8.6)
subtype1	9	71.0 (8.9)
subtype2	14	68.8 (9.6)
subtype4	8	70.2 (7.3)

Figure S10. Get High-res Image Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #2: 'AGE'

Clustering Approach #6: 'MIRSEQ CHIERARCHICAL'

Table S16. Description of clustering approach #6: 'MIRSEQ CHIERARCHICAL'

Cluster Labels	1	2	3
Number of samples	1	15	17

'MIRSEQ CHIERARCHICAL' versus 'Time to Death'

P value = 0.84 (logrank test), Q value = 1

Table S17. Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	32	19	4.6 - 102.4 (17.2)
subtype2	15	8	4.6 - 102.4 (14.7)
subtype3	17	11	4.9 - 85.3 (24.0)

Figure S11. Get High-res Image Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CHIERARCHICAL' versus 'AGE'

P value = 0.151 (t-test), Q value = 1

Table S18. Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	32	69.9 (8.4)
subtype2	15	72.1 (7.4)
subtype3	17	67.9 (8.9)

Figure S12. Get High-res Image Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'AGE'

Clustering Approach #7: 'MIRseq Mature CNMF subtypes'

Table S19. Description of clustering approach #7: 'MIRseq Mature CNMF subtypes'

Cluster Labels	1	2	3	4
Number of samples	10	15	2	6

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

P value = 0.69 (logrank test), Q value = 1

Table S20. Clustering Approach #7: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	31	19	3.7 - 102.4 (16.3)
subtype1	10	7	3.7 - 102.4 (10.9)
subtype2	15	10	4.9 - 85.3 (24.0)
subtype4	6	2	5.4 - 31.2 (14.8)

Figure S13. Get High-res Image Clustering Approach #7: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature CNMF subtypes' versus 'AGE'

P value = 0.551 (ANOVA), Q value = 1

Table S21. Clustering Approach #7: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	31	69.5 (8.6)
subtype1	10	68.8 (9.0)
subtype2	15	68.5 (9.3)
subtype4	6	73.0 (6.2)

Figure S14. Get High-res Image Clustering Approach #7: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'AGE'

Clustering Approach #8: 'MIRseq Mature cHierClus subtypes'

Table S22. Description of clustering approach #8: 'MIRseq Mature cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	1	16	16

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

P value = 0.96 (logrank test), Q value = 1

Table S23. Clustering Approach #8: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	32	19	4.6 - 102.4 (17.2)
subtype2	16	11	4.9 - 85.3 (25.3)
subtype3	16	8	4.6 - 102.4 (14.6)

Figure S15. Get High-res Image Clustering Approach #8: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature cHierClus subtypes' versus 'AGE'

P value = 0.212 (t-test), Q value = 1

Table S24. Clustering Approach #8: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	32	69.9 (8.4)
subtype2	16	68.0 (9.2)
subtype3	16	71.8 (7.3)

Figure S16. Get High-res Image Clustering Approach #8: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'AGE'

Methods & Data

Input

Cluster data file = UCS-TP.mergedcluster.txt
Clinical data file = UCS-TP.merged_data.txt
Number of patients = 33
Number of clustering approaches = 8
Number of selected clinical features = 2
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

Student's t-test analysis

For continuous numerical clinical features, two-tailed Student's t test with unequal variance (Lehmann and Romano 2005) was applied to compare the clinical values between two tumor subtypes using 't.test' function in R

Q value calculation

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

References

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

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

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

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

[5] Lehmann and Romano, Testing Statistical Hypotheses (3E ed.), New York: Springer. ISBN 0387988645 (2005)

[6] 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