Correlation between aggregated molecular cancer subtypes and selected clinical features

Acute Myeloid Leukemia (Primary blood derived cancer - Peripheral blood)

17 October 2014 | analyses__2014_10_17

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 (2014): Correlation between aggregated molecular cancer subtypes and selected clinical features. Broad Institute of MIT and Harvard. doi:10.7908/C1D21WGW

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 6 different clustering approaches and 5 clinical features across 200 patients, 5 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 'Time to Death'.
5 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. These subtypes correlate to 'Time to Death' and 'AGE'.
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 6 subtypes that correlate to 'Time to Death' and 'AGE'.
4 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.

Results

Overview of the results

Table 1. Get Full Table Overview of the association between subtypes identified by 6 different clustering approaches and 5 clinical features. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, 5 significant findings detected.


Clinical Features	Time to Death	AGE	GENDER	RACE	ETHNICITY
Statistical Tests	logrank test	Kruskal-Wallis (anova)	Fisher's exact test	Fisher's exact test	Fisher's exact test
Copy Number Ratio CNMF subtypes	0.00346 (0.0899)	0.0454 (1.00)	0.075 (1.00)	1 (1.00)	0.258 (1.00)
METHLYATION CNMF	9.57e-05 (0.00268)	2.57e-09 (7.72e-08)	0.32 (1.00)	0.677 (1.00)	1 (1.00)
RNAseq CNMF subtypes	0.502 (1.00)	0.24 (1.00)	0.101 (1.00)	0.0366 (0.877)	0.583 (1.00)
RNAseq cHierClus subtypes	0.000903 (0.0244)	3.26e-05 (0.000947)	0.414 (1.00)	0.146 (1.00)	0.0819 (1.00)
MIRSEQ CNMF	0.0943 (1.00)	0.0105 (0.262)	0.923 (1.00)	0.858 (1.00)	1 (1.00)
MIRSEQ CHIERARCHICAL	0.194 (1.00)	0.457 (1.00)	0.391 (1.00)	0.956 (1.00)	1 (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	4
Number of samples	148	14	27	2

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

P value = 0.00346 (logrank test), Q value = 0.09

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

	nPatients	nDeath	Duration Range (Median), Month
ALL	166	105	0.9 - 94.1 (12.0)
subtype1	131	77	0.9 - 94.1 (13.9)
subtype2	11	8	0.9 - 42.0 (10.0)
subtype3	24	20	1.0 - 73.0 (9.5)

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 = 0.0454 (Kruskal-Wallis (anova)), Q value = 1

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

	nPatients	Mean (Std.Dev)
ALL	189	55.1 (16.1)
subtype1	148	53.7 (16.0)
subtype2	14	62.4 (13.1)
subtype3	27	58.7 (16.6)

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 = 0.075 (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	86	103
subtype1	73	75
subtype2	6	8
subtype3	7	20

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 'RACE'

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

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

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	14	171
subtype1	2	11	133
subtype2	0	1	13
subtype3	0	2	25

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

'Copy Number Ratio CNMF subtypes' versus 'ETHNICITY'

P value = 0.258 (Fisher's exact test), Q value = 1

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

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	3	183
subtype1	2	143
subtype2	1	13
subtype3	0	27

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

Clustering Approach #2: 'METHLYATION CNMF'

Table S7. Description of clustering approach #2: 'METHLYATION CNMF'

Cluster Labels	1	2	3	4	5
Number of samples	48	45	65	14	22

'METHLYATION CNMF' versus 'Time to Death'

P value = 9.57e-05 (logrank test), Q value = 0.0027

Table S8. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	169	106	0.9 - 94.1 (12.0)
subtype1	42	30	1.0 - 69.0 (8.1)
subtype2	42	13	0.9 - 94.1 (21.6)
subtype3	55	43	0.9 - 56.1 (12.0)
subtype4	13	7	1.0 - 42.0 (13.0)
subtype5	17	13	4.0 - 73.0 (12.0)

Figure S6. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'AGE'

P value = 2.57e-09 (Kruskal-Wallis (anova)), Q value = 7.7e-08

Table S9. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	194	55.1 (16.0)
subtype1	48	55.9 (14.1)
subtype2	45	45.6 (15.5)
subtype3	65	63.1 (12.9)
subtype4	14	63.5 (11.2)
subtype5	22	43.9 (16.0)

Figure S7. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

'METHLYATION CNMF' versus 'GENDER'

P value = 0.32 (Fisher's exact test), Q value = 1

Table S10. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	89	105
subtype1	24	24
subtype2	24	21
subtype3	26	39
subtype4	8	6
subtype5	7	15

Figure S8. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'GENDER'

'METHLYATION CNMF' versus 'RACE'

P value = 0.677 (Fisher's exact test), Q value = 1

Table S11. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	13	177
subtype1	1	2	45
subtype2	1	2	41
subtype3	0	6	59
subtype4	0	2	12
subtype5	0	1	20

Figure S9. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'RACE'

'METHLYATION CNMF' versus 'ETHNICITY'

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

Table S12. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	1	190
subtype1	0	48
subtype2	0	44
subtype3	1	63
subtype4	0	14
subtype5	0	21

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

Clustering Approach #3: 'RNAseq CNMF subtypes'

Table S13. Description of clustering approach #3: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	74	54	45

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.502 (logrank test), Q value = 1

Table S14. Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	151	94	0.9 - 94.1 (12.0)
subtype1	65	39	1.0 - 94.1 (16.1)
subtype2	48	31	0.9 - 75.1 (10.5)
subtype3	38	24	0.9 - 62.0 (12.0)

Figure S11. Get High-res Image Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'AGE'

P value = 0.24 (Kruskal-Wallis (anova)), Q value = 1

Table S15. Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	173	55.3 (16.1)
subtype1	74	54.4 (17.3)
subtype2	54	58.4 (13.9)
subtype3	45	52.9 (16.4)

Figure S12. Get High-res Image Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 0.101 (Fisher's exact test), Q value = 1

Table S16. Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	80	93
subtype1	31	43
subtype2	22	32
subtype3	27	18

Figure S13. Get High-res Image Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'GENDER'

'RNAseq CNMF subtypes' versus 'RACE'

P value = 0.0366 (Fisher's exact test), Q value = 0.88

Table S17. Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	13	156
subtype1	1	3	68
subtype2	1	2	51
subtype3	0	8	37

Figure S14. Get High-res Image Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'RACE'

'RNAseq CNMF subtypes' versus 'ETHNICITY'

P value = 0.583 (Fisher's exact test), Q value = 1

Table S18. Clustering Approach #3: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	1	169
subtype1	0	71
subtype2	1	53
subtype3	0	45

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

Clustering Approach #4: 'RNAseq cHierClus subtypes'

Table S19. Description of clustering approach #4: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3	4	5	6
Number of samples	16	14	15	58	26	44

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.000903 (logrank test), Q value = 0.024

Table S20. Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	151	94	0.9 - 94.1 (12.0)
subtype1	15	4	0.9 - 62.0 (46.1)
subtype2	14	5	0.9 - 75.1 (20.0)
subtype3	13	6	6.0 - 94.1 (19.0)
subtype4	50	38	1.0 - 73.0 (11.5)
subtype5	23	13	1.0 - 62.0 (10.0)
subtype6	36	28	1.0 - 69.0 (9.5)

Figure S16. Get High-res Image Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'AGE'

P value = 3.26e-05 (Kruskal-Wallis (anova)), Q value = 0.00095

Table S21. Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	173	55.3 (16.1)
subtype1	16	47.8 (14.9)
subtype2	14	57.7 (14.6)
subtype3	15	36.6 (13.9)
subtype4	58	59.8 (15.5)
subtype5	26	53.3 (16.7)
subtype6	44	58.7 (13.0)

Figure S17. Get High-res Image Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 0.414 (Fisher's exact test), Q value = 1

Table S22. Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	80	93
subtype1	9	7
subtype2	6	8
subtype3	6	9
subtype4	22	36
subtype5	16	10
subtype6	21	23

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

'RNAseq cHierClus subtypes' versus 'RACE'

P value = 0.146 (Fisher's exact test), Q value = 1

Table S23. Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	13	156
subtype1	0	2	13
subtype2	0	0	14
subtype3	0	0	15
subtype4	1	3	53
subtype5	0	6	20
subtype6	1	2	41

Figure S19. Get High-res Image Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'RACE'

'RNAseq cHierClus subtypes' versus 'ETHNICITY'

P value = 0.0819 (Fisher's exact test), Q value = 1

Table S24. Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	1	169
subtype1	0	15
subtype2	1	13
subtype3	0	15
subtype4	0	56
subtype5	0	26
subtype6	0	44

Figure S20. Get High-res Image Clustering Approach #4: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'ETHNICITY'

Clustering Approach #5: 'MIRSEQ CNMF'

Table S25. Description of clustering approach #5: 'MIRSEQ CNMF'

Cluster Labels	1	2	3	4
Number of samples	60	38	43	47

'MIRSEQ CNMF' versus 'Time to Death'

P value = 0.0943 (logrank test), Q value = 1

Table S26. Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	164	102	0.9 - 94.1 (12.0)
subtype1	53	38	1.0 - 69.0 (9.0)
subtype2	34	18	0.9 - 62.0 (14.5)
subtype3	35	23	1.0 - 94.1 (12.0)
subtype4	42	23	0.9 - 73.0 (15.5)

Figure S21. Get High-res Image Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CNMF' versus 'AGE'

P value = 0.0105 (Kruskal-Wallis (anova)), Q value = 0.26

Table S27. Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	188	54.9 (16.2)
subtype1	60	57.4 (14.3)
subtype2	38	56.6 (14.9)
subtype3	43	57.4 (18.0)
subtype4	47	48.0 (16.2)

Figure S22. Get High-res Image Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #2: 'AGE'

'MIRSEQ CNMF' versus 'GENDER'

P value = 0.923 (Fisher's exact test), Q value = 1

Table S28. Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	87	101
subtype1	26	34
subtype2	17	21
subtype3	21	22
subtype4	23	24

Figure S23. Get High-res Image Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #3: 'GENDER'

'MIRSEQ CNMF' versus 'RACE'

P value = 0.858 (Fisher's exact test), Q value = 1

Table S29. Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #4: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	13	171
subtype1	1	4	55
subtype2	0	2	36
subtype3	0	2	40
subtype4	1	5	40

Figure S24. Get High-res Image Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #4: 'RACE'

'MIRSEQ CNMF' versus 'ETHNICITY'

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

Table S30. Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #5: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	1	184
subtype1	1	59
subtype2	0	38
subtype3	0	41
subtype4	0	46

Figure S25. Get High-res Image Clustering Approach #5: 'MIRSEQ CNMF' versus Clinical Feature #5: 'ETHNICITY'

Clustering Approach #6: 'MIRSEQ CHIERARCHICAL'

Table S31. Description of clustering approach #6: 'MIRSEQ CHIERARCHICAL'

Cluster Labels	1	2	3
Number of samples	70	35	83

'MIRSEQ CHIERARCHICAL' versus 'Time to Death'

P value = 0.194 (logrank test), Q value = 1

Table S32. Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	164	102	0.9 - 94.1 (12.0)
subtype1	62	42	1.0 - 69.0 (9.5)
subtype2	30	16	0.9 - 62.0 (18.0)
subtype3	72	44	0.9 - 94.1 (13.5)

Figure S26. Get High-res Image Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CHIERARCHICAL' versus 'AGE'

P value = 0.457 (Kruskal-Wallis (anova)), Q value = 1

Table S33. Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	188	54.9 (16.2)
subtype1	70	55.8 (14.8)
subtype2	35	57.3 (15.2)
subtype3	83	53.1 (17.6)

Figure S27. Get High-res Image Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'AGE'

'MIRSEQ CHIERARCHICAL' versus 'GENDER'

P value = 0.391 (Fisher's exact test), Q value = 1

Table S34. Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'GENDER'

nPatients	FEMALE	MALE
ALL	87	101
subtype1	36	34
subtype2	13	22
subtype3	38	45

Figure S28. Get High-res Image Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'GENDER'

'MIRSEQ CHIERARCHICAL' versus 'RACE'

P value = 0.956 (Fisher's exact test), Q value = 1

Table S35. Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	13	171
subtype1	1	6	63
subtype2	0	2	33
subtype3	1	5	75

Figure S29. Get High-res Image Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'RACE'

'MIRSEQ CHIERARCHICAL' versus 'ETHNICITY'

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

Table S36. Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	1	184
subtype1	0	70
subtype2	0	35
subtype3	1	79

Figure S30. Get High-res Image Clustering Approach #6: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'ETHNICITY'

Methods & Data

Input

Cluster data file = LAML-TB.mergedcluster.txt
Clinical data file = LAML-TB.merged_data.txt
Number of patients = 200
Number of clustering approaches = 6
Number of selected clinical features = 5
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

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] 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] 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