Correlation between aggregated molecular cancer subtypes and selected clinical features

Esophageal Carcinoma (Primary solid tumor)

23 September 2013 | analyses__2013_09_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/C1NK3CB0

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 2 different clustering approaches and 7 clinical features across 19 patients, one 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 correlate to 'AGE'.
3 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. 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 2 different clustering approaches and 7 clinical features. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, one significant finding detected.


Clinical Features	Statistical Tests	Copy Number Ratio CNMF subtypes	METHLYATION CNMF
Time to Death	logrank test	0.0543 (0.706)	0.895 (1.00)
AGE	ANOVA	0.00223 (0.0313)	0.276 (1.00)
NEOPLASM DISEASESTAGE	Chi-square test	0.8 (1.00)	0.373 (1.00)
PATHOLOGY T STAGE	Chi-square test	0.59 (1.00)	0.4 (1.00)
PATHOLOGY N STAGE	Fisher's exact test	0.325 (1.00)	0.281 (1.00)
GENDER	Fisher's exact test	0.74 (1.00)	1 (1.00)
NUMBERPACKYEARSSMOKED	ANOVA	0.761 (1.00)	0.515 (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	7	3	9

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

P value = 0.0543 (logrank test), Q value = 0.71

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

	nPatients	nDeath	Duration Range (Median), Month
ALL	15	5	0.0 - 30.7 (3.5)
subtype1	4	0	0.0 - 4.1 (0.1)
subtype2	3	2	0.3 - 7.1 (1.4)
subtype3	8	3	3.2 - 30.7 (5.6)

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 = 0.00223 (ANOVA), Q value = 0.031

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

	nPatients	Mean (Std.Dev)
ALL	19	66.4 (11.4)
subtype1	7	56.3 (5.1)
subtype2	3	67.3 (10.0)
subtype3	9	74.0 (9.5)

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 'NEOPLASM.DISEASESTAGE'

P value = 0.8 (Chi-square test), Q value = 1

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

nPatients	STAGE IA	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IIIC
ALL	2	3	7	4	2	1
subtype1	1	2	2	2	0	0
subtype2	0	0	1	1	1	0
subtype3	1	1	4	1	1	1

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

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY.T.STAGE'

P value = 0.59 (Chi-square test), Q value = 1

Table S5. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY.T.STAGE'

nPatients	T1	T2	T3+T4
ALL	4	7	8
subtype1	1	3	3
subtype2	0	2	1
subtype3	3	2	4

Figure S4. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY.T.STAGE'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY.N.STAGE'

P value = 0.325 (Fisher's exact test), Q value = 1

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

nPatients	N0	N1+N2+N3
ALL	7	12
subtype1	4	3
subtype2	0	3
subtype3	3	6

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

'Copy Number Ratio CNMF subtypes' versus 'GENDER'

P value = 0.74 (Fisher's exact test), Q value = 1

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

nPatients	FEMALE	MALE
ALL	3	16
subtype1	1	6
subtype2	1	2
subtype3	1	8

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

'Copy Number Ratio CNMF subtypes' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.761 (ANOVA), Q value = 1

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

	nPatients	Mean (Std.Dev)
ALL	12	29.3 (13.8)
subtype1	4	29.4 (10.9)
subtype2	2	37.5 (3.5)
subtype3	6	26.5 (17.6)

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

Clustering Approach #2: 'METHLYATION CNMF'

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

Cluster Labels	1	2	3
Number of samples	6	4	9

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.895 (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	15	5	0.0 - 30.7 (3.5)
subtype1	3	0	0.1 - 4.1 (0.1)
subtype2	3	1	0.0 - 30.7 (1.4)
subtype3	9	4	0.3 - 29.0 (3.7)

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

'METHLYATION CNMF' versus 'AGE'

P value = 0.276 (ANOVA), Q value = 1

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

	nPatients	Mean (Std.Dev)
ALL	19	66.4 (11.4)
subtype1	6	60.8 (6.3)
subtype2	4	65.5 (18.0)
subtype3	9	70.6 (10.0)

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

'METHLYATION CNMF' versus 'NEOPLASM.DISEASESTAGE'

P value = 0.373 (Chi-square test), Q value = 1

Table S12. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'NEOPLASM.DISEASESTAGE'

nPatients	STAGE IA	STAGE IIA	STAGE IIB	STAGE IIIA	STAGE IIIB	STAGE IIIC
ALL	2	3	7	4	2	1
subtype1	1	2	1	2	0	0
subtype2	0	0	2	2	0	0
subtype3	1	1	4	0	2	1

Figure S10. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'NEOPLASM.DISEASESTAGE'

'METHLYATION CNMF' versus 'PATHOLOGY.T.STAGE'

P value = 0.4 (Chi-square test), Q value = 1

Table S13. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY.T.STAGE'

nPatients	T1	T2	T3+T4
ALL	4	7	8
subtype1	1	2	3
subtype2	0	3	1
subtype3	3	2	4

Figure S11. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY.T.STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY.N.STAGE'

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

Table S14. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'PATHOLOGY.N.STAGE'

nPatients	N0	N1+N2+N3
ALL	7	12
subtype1	4	2
subtype2	1	3
subtype3	2	7

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

'METHLYATION CNMF' versus 'GENDER'

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

Table S15. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'GENDER'

nPatients	FEMALE	MALE
ALL	3	16
subtype1	1	5
subtype2	1	3
subtype3	1	8

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

'METHLYATION CNMF' versus 'NUMBERPACKYEARSSMOKED'

P value = 0.515 (ANOVA), Q value = 1

Table S16. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'NUMBERPACKYEARSSMOKED'

	nPatients	Mean (Std.Dev)
ALL	12	29.3 (13.8)
subtype1	3	36.7 (15.3)
subtype2	3	30.8 (9.5)
subtype3	6	24.9 (15.3)

Figure S14. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'NUMBERPACKYEARSSMOKED'

Methods & Data

Input

Cluster data file = ESCA-TP.mergedcluster.txt
Clinical data file = ESCA-TP.clin.merged.picked.txt
Number of patients = 19
Number of clustering approaches = 2
Number of selected clinical features = 7
Exclude small clusters that include fewer than K patients, K = 3

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

In addition to the links below, the full results of the analysis summarized in this report can also be downloaded programmatically using firehose_get, or interactively from either the Broad GDAC website or TCGA Data Coordination Center Portal.

References

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

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

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

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

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

Made with Nozzle