Colon/Rectal Adenocarcinoma: Correlation between molecular cancer subtypes and selected clinical features

Maintained by TCGA GDAC Team (Broad Institute/Dana-Farber Cancer Institute/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 7 different clustering approaches and 10 clinical features across 585 patients, 12 significant findings detected with P value < 0.05.

CNMF clustering analysis on array-based mRNA expression data identified 4 subtypes that correlate to 'PRIMARY.SITE.OF.DISEASE', 'HISTOLOGICAL.TYPE', 'PATHOLOGY.N', and 'PATHOLOGICSPREAD(M)'.
Consensus hierarchical clustering analysis on array-based mRNA expression data identified 3 subtypes that correlate to 'PRIMARY.SITE.OF.DISEASE' and 'HISTOLOGICAL.TYPE'.
3 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. These subtypes correlate to 'PRIMARY.SITE.OF.DISEASE' and 'HISTOLOGICAL.TYPE'.
CNMF clustering analysis on sequencing-based mRNA expression data identified 4 subtypes that correlate to 'HISTOLOGICAL.TYPE'.
Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 4 subtypes that correlate to 'HISTOLOGICAL.TYPE'.
CNMF clustering analysis on sequencing-based miR expression data identified 3 subtypes that correlate to 'PATHOLOGY.N' and 'NEOADJUVANT.THERAPY'.
Consensus hierarchical clustering analysis on sequencing-based miR expression data identified 2 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 7 different clustering approaches and 10 clinical features. Shown in the table are P values from statistical tests. Thresholded by P value < 0.05, 12 significant findings detected.


Clinical Features	Statistical Tests	mRNA CNMF subtypes	mRNA cHierClus subtypes	METHLYATION CNMF	RNAseq CNMF subtypes	RNAseq cHierClus subtypes	MIRseq CNMF subtypes	MIRseq cHierClus subtypes
Time to Death	logrank test	0.145	0.805	0.846	0.12	0.176	0.0895	0.738
AGE	t-test	0.596	0.148	0.135	0.416	0.884	0.82	0.472
PRIMARY SITE OF DISEASE	Fisher's exact test	0.0126	0.035	0.00011	0.111	0.265	0.965	0.76
GENDER	Fisher's exact test	0.072	0.796	0.162	0.869	0.924	0.653	0.42
HISTOLOGICAL TYPE	Chi-square test	6.39e-07	3.6e-08	1.72e-05	0.0132	0.00215	0.942	0.513
PATHOLOGY T	Chi-square test	0.184	0.866	0.177	0.439	0.71	0.331	0.208
PATHOLOGY N	Chi-square test	0.00595	0.65	0.529	0.872	0.969	0.0292	0.946
PATHOLOGICSPREAD(M)	Chi-square test	0.00173	0.498	0.578	0.28	0.852	0.1	0.313
RADIATIONS RADIATION REGIMENINDICATION	Fisher's exact test	0.393	0.496	0.0864	0.41	1	0.553	0.314
NEOADJUVANT THERAPY	Fisher's exact test	0.127	0.712	0.626	0.41	1	0.0117	0.237

Clustering Approach #1: 'mRNA CNMF subtypes'

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

Cluster Labels	1	2	3	4
Number of samples	63	48	73	40

'mRNA CNMF subtypes' versus 'Time to Death'

P value = 0.145 (logrank test)

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

	nPatients	nDeath	Duration Range (Median), Month
ALL	114	15	0.9 - 52.0 (5.5)
subtype1	36	9	1.0 - 52.0 (13.0)
subtype2	26	2	1.0 - 30.0 (1.0)
subtype3	28	2	0.9 - 49.9 (1.0)
subtype4	24	2	0.9 - 52.0 (12.5)

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

'mRNA CNMF subtypes' versus 'AGE'

P value = 0.596 (ANOVA)

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

	nPatients	Mean (Std.Dev)
ALL	224	69.3 (11.5)
subtype1	63	67.7 (9.7)
subtype2	48	70.3 (11.8)
subtype3	73	69.6 (12.3)
subtype4	40	70.2 (12.2)

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

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

P value = 0.0126 (Fisher's exact test)

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

nPatients	COLON	RECTUM
ALL	154	68
subtype1	37	26
subtype2	41	7
subtype3	46	25
subtype4	30	10

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

'mRNA CNMF subtypes' versus 'GENDER'

P value = 0.072 (Fisher's exact test)

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

nPatients	FEMALE	MALE
ALL	117	107
subtype1	30	33
subtype2	19	29
subtype3	42	31
subtype4	26	14

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

'mRNA CNMF subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 6.39e-07 (Chi-square test)

Table S6. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients	COLON ADENOCARCINOMA	COLON MUCINOUS ADENOCARCINOMA	RECTAL ADENOCARCINOMA	RECTAL MUCINOUS ADENOCARCINOMA
ALL	128	24	58	7
subtype1	36	1	26	0
subtype2	27	13	4	2
subtype3	45	1	21	2
subtype4	20	9	7	3

Figure S5. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'mRNA CNMF subtypes' versus 'PATHOLOGY.T'

P value = 0.184 (Chi-square test)

Table S7. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients	T0+T1	T2	T3	T4
ALL	9	46	150	17
subtype1	1	12	42	8
subtype2	1	9	32	4
subtype3	5	16	52	0
subtype4	2	9	24	5

Figure S6. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

'mRNA CNMF subtypes' versus 'PATHOLOGY.N'

P value = 0.00595 (Chi-square test)

Table S8. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients	N0	N1	N2
ALL	137	43	44
subtype1	28	18	17
subtype2	30	6	12
subtype3	47	12	14
subtype4	32	7	1

Figure S7. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

'mRNA CNMF subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 0.00173 (Chi-square test)

Table S9. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients	M0	M1	M1A
ALL	186	34	1
subtype1	44	19	0
subtype2	41	6	0
subtype3	64	8	0
subtype4	37	1	1

Figure S8. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'mRNA CNMF subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.393 (Fisher's exact test)

Table S10. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients	NO	YES
ALL	223	1
subtype1	63	0
subtype2	47	1
subtype3	73	0
subtype4	40	0

Figure S9. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'mRNA CNMF subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 0.127 (Fisher's exact test)

Table S11. Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

nPatients	NO	YES
ALL	216	8
subtype1	60	3
subtype2	44	4
subtype3	72	1
subtype4	40	0

Figure S10. Get High-res Image Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

Clustering Approach #2: 'mRNA cHierClus subtypes'

Table S12. Get Full Table Description of clustering approach #2: 'mRNA cHierClus subtypes'

Cluster Labels	1	2	3
Number of samples	47	64	113

'mRNA cHierClus subtypes' versus 'Time to Death'

P value = 0.805 (logrank test)

Table S13. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	114	15	0.9 - 52.0 (5.5)
subtype1	12	1	0.9 - 17.0 (1.0)
subtype2	35	3	1.0 - 41.0 (1.0)
subtype3	67	11	0.9 - 52.0 (12.0)

Figure S11. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'mRNA cHierClus subtypes' versus 'AGE'

P value = 0.148 (ANOVA)

Table S14. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	224	69.3 (11.5)
subtype1	47	67.5 (13.0)
subtype2	64	71.6 (11.3)
subtype3	113	68.8 (10.8)

Figure S12. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #2: 'AGE'

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

P value = 0.035 (Fisher's exact test)

Table S15. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	COLON	RECTUM
ALL	154	68
subtype1	34	12
subtype2	51	13
subtype3	69	43

Figure S13. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'mRNA cHierClus subtypes' versus 'GENDER'

P value = 0.796 (Fisher's exact test)

Table S16. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	117	107
subtype1	25	22
subtype2	31	33
subtype3	61	52

Figure S14. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

'mRNA cHierClus subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 3.6e-08 (Chi-square test)

Table S17. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients	COLON ADENOCARCINOMA	COLON MUCINOUS ADENOCARCINOMA	RECTAL ADENOCARCINOMA	RECTAL MUCINOUS ADENOCARCINOMA
ALL	128	24	58	7
subtype1	33	1	11	1
subtype2	29	20	9	3
subtype3	66	3	38	3

Figure S15. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'mRNA cHierClus subtypes' versus 'PATHOLOGY.T'

P value = 0.866 (Chi-square test)

Table S18. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients	T0+T1	T2	T3	T4
ALL	9	46	150	17
subtype1	1	9	35	2
subtype2	2	13	41	6
subtype3	6	24	74	9

Figure S16. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

'mRNA cHierClus subtypes' versus 'PATHOLOGY.N'

P value = 0.65 (Chi-square test)

Table S19. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients	N0	N1	N2
ALL	137	43	44
subtype1	28	11	8
subtype2	43	11	10
subtype3	66	21	26

Figure S17. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

'mRNA cHierClus subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 0.498 (Chi-square test)

Table S20. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients	M0	M1	M1A
ALL	186	34	1
subtype1	38	8	0
subtype2	57	6	0
subtype3	91	20	1

Figure S18. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'mRNA cHierClus subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.496 (Fisher's exact test)

Table S21. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients	NO	YES
ALL	223	1
subtype1	47	0
subtype2	63	1
subtype3	113	0

Figure S19. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'mRNA cHierClus subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 0.712 (Fisher's exact test)

Table S22. Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

nPatients	NO	YES
ALL	216	8
subtype1	45	2
subtype2	61	3
subtype3	110	3

Figure S20. Get High-res Image Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

Clustering Approach #3: 'METHLYATION CNMF'

Table S23. Get Full Table Description of clustering approach #3: 'METHLYATION CNMF'

Cluster Labels	1	2	3
Number of samples	142	97	110

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.846 (logrank test)

Table S24. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	324	44	0.1 - 135.5 (7.0)
subtype1	132	19	0.1 - 135.5 (7.2)
subtype2	90	10	0.1 - 102.4 (6.0)
subtype3	102	15	0.1 - 129.1 (7.2)

Figure S21. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'AGE'

P value = 0.135 (ANOVA)

Table S25. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	348	64.8 (13.0)
subtype1	142	65.1 (12.7)
subtype2	96	62.7 (12.8)
subtype3	110	66.3 (13.5)

Figure S22. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #2: 'AGE'

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

P value = 0.00011 (Fisher's exact test)

Table S26. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	COLON	RECTUM
ALL	254	93
subtype1	92	50
subtype2	66	29
subtype3	96	14

Figure S23. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'METHLYATION CNMF' versus 'GENDER'

P value = 0.162 (Fisher's exact test)

Table S27. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	193	156
subtype1	81	61
subtype2	59	38
subtype3	53	57

Figure S24. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #4: 'GENDER'

'METHLYATION CNMF' versus 'HISTOLOGICAL.TYPE'

P value = 1.72e-05 (Chi-square test)

Table S28. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients	COLON ADENOCARCINOMA	COLON MUCINOUS ADENOCARCINOMA	RECTAL ADENOCARCINOMA	RECTAL MUCINOUS ADENOCARCINOMA
ALL	224	30	87	6
subtype1	89	3	48	2
subtype2	52	14	28	1
subtype3	83	13	11	3

Figure S25. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'METHLYATION CNMF' versus 'PATHOLOGY.T'

P value = 0.177 (Chi-square test)

Table S29. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients	T0+T1	T2	T3	T4
ALL	11	51	243	42
subtype1	5	21	100	16
subtype2	3	8	68	17
subtype3	3	22	75	9

Figure S26. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #6: 'PATHOLOGY.T'

'METHLYATION CNMF' versus 'PATHOLOGY.N'

P value = 0.529 (Chi-square test)

Table S30. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients	N0	N1	N2
ALL	193	93	58
subtype1	73	42	24
subtype2	51	27	17
subtype3	69	24	17

Figure S27. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #7: 'PATHOLOGY.N'

'METHLYATION CNMF' versus 'PATHOLOGICSPREAD(M)'

P value = 0.578 (Chi-square test)

Table S31. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients	M0	M1	M1A	M1B	MX
ALL	246	37	8	1	50
subtype1	96	20	4	0	21
subtype2	69	8	3	0	14
subtype3	81	9	1	1	15

Figure S28. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'METHLYATION CNMF' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.0864 (Fisher's exact test)

Table S32. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients	NO	YES
ALL	342	7
subtype1	139	3
subtype2	93	4
subtype3	110	0

Figure S29. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'METHLYATION CNMF' versus 'NEOADJUVANT.THERAPY'

P value = 0.626 (Fisher's exact test)

Table S33. Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

nPatients	NO	YES
ALL	280	69
subtype1	114	28
subtype2	75	22
subtype3	91	19

Figure S30. Get High-res Image Clustering Approach #3: 'METHLYATION CNMF' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

Clustering Approach #4: 'RNAseq CNMF subtypes'

Table S34. Get Full Table Description of clustering approach #4: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3	4
Number of samples	17	17	27	22

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.12 (logrank test)

Table S35. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	46	4	0.9 - 72.1 (8.2)
subtype1	6	1	1.0 - 64.0 (11.6)
subtype2	8	0	1.0 - 17.0 (1.0)
subtype3	16	0	0.9 - 61.9 (8.5)
subtype4	16	3	0.9 - 72.1 (7.0)

Figure S31. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'AGE'

P value = 0.416 (ANOVA)

Table S36. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	83	67.6 (10.5)
subtype1	17	64.8 (9.6)
subtype2	17	66.1 (9.8)
subtype3	27	68.2 (12.3)
subtype4	22	70.1 (9.3)

Figure S32. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

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

P value = 0.111 (Fisher's exact test)

Table S37. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	COLON	RECTUM
ALL	11	71
subtype1	3	14
subtype2	3	13
subtype3	5	22
subtype4	0	22

Figure S33. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 0.869 (Fisher's exact test)

Table S38. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	39	44
subtype1	9	8
subtype2	8	9
subtype3	11	16
subtype4	11	11

Figure S34. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'GENDER'

'RNAseq CNMF subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 0.0132 (Chi-square test)

Table S39. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients	COLON ADENOCARCINOMA	RECTAL ADENOCARCINOMA	RECTAL MUCINOUS ADENOCARCINOMA
ALL	11	60	8
subtype1	3	11	2
subtype2	3	13	0
subtype3	5	14	6
subtype4	0	22	0

Figure S35. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY.T'

P value = 0.439 (Chi-square test)

Table S40. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients	T0+T1	T2	T3	T4
ALL	5	19	54	5
subtype1	0	2	14	1
subtype2	1	4	12	0
subtype3	3	6	17	1
subtype4	1	7	11	3

Figure S36. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

'RNAseq CNMF subtypes' versus 'PATHOLOGY.N'

P value = 0.872 (Chi-square test)

Table S41. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients	N0	N1	N2
ALL	52	17	14
subtype1	9	4	4
subtype2	12	3	2
subtype3	18	6	3
subtype4	13	4	5

Figure S37. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

'RNAseq CNMF subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 0.28 (Chi-square test)

Table S42. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients	M0	M1	M1A
ALL	68	13	1
subtype1	13	4	0
subtype2	16	1	0
subtype3	23	2	1
subtype4	16	6	0

Figure S38. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'RNAseq CNMF subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.41 (Fisher's exact test)

Table S43. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients	NO	YES
ALL	82	1
subtype1	16	1
subtype2	17	0
subtype3	27	0
subtype4	22	0

Figure S39. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'RNAseq CNMF subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 0.41 (Fisher's exact test)

Table S44. Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

nPatients	NO	YES
ALL	82	1
subtype1	16	1
subtype2	17	0
subtype3	27	0
subtype4	22	0

Figure S40. Get High-res Image Clustering Approach #4: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

Clustering Approach #5: 'RNAseq cHierClus subtypes'

Table S45. Get Full Table Description of clustering approach #5: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3	4
Number of samples	17	16	28	22

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.176 (logrank test)

Table S46. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	46	4	0.9 - 72.1 (8.2)
subtype1	9	1	0.9 - 38.9 (2.0)
subtype2	11	1	1.0 - 52.0 (12.0)
subtype3	14	0	0.9 - 61.9 (8.5)
subtype4	12	2	1.0 - 72.1 (1.5)

Figure S41. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'AGE'

P value = 0.884 (ANOVA)

Table S47. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	83	67.6 (10.5)
subtype1	17	69.2 (9.1)
subtype2	16	66.5 (9.4)
subtype3	28	67.7 (13.0)
subtype4	22	67.0 (9.2)

Figure S42. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

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

P value = 0.265 (Fisher's exact test)

Table S48. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	COLON	RECTUM
ALL	11	71
subtype1	1	15
subtype2	1	15
subtype3	7	21
subtype4	2	20

Figure S43. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 0.924 (Fisher's exact test)

Table S49. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	39	44
subtype1	8	9
subtype2	8	8
subtype3	14	14
subtype4	9	13

Figure S44. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

'RNAseq cHierClus subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 0.00215 (Chi-square test)

Table S50. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients	COLON ADENOCARCINOMA	RECTAL ADENOCARCINOMA	RECTAL MUCINOUS ADENOCARCINOMA
ALL	11	60	8
subtype1	1	13	1
subtype2	1	15	0
subtype3	7	12	7
subtype4	2	20	0

Figure S45. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY.T'

P value = 0.71 (Chi-square test)

Table S51. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients	T0+T1	T2	T3	T4
ALL	5	19	54	5
subtype1	1	4	12	0
subtype2	2	4	9	1
subtype3	2	6	19	1
subtype4	0	5	14	3

Figure S46. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY.N'

P value = 0.969 (Chi-square test)

Table S52. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients	N0	N1	N2
ALL	52	17	14
subtype1	10	4	3
subtype2	10	4	2
subtype3	19	5	4
subtype4	13	4	5

Figure S47. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

'RNAseq cHierClus subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 0.852 (Chi-square test)

Table S53. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients	M0	M1	M1A
ALL	68	13	1
subtype1	14	3	0
subtype2	13	3	0
subtype3	23	3	1
subtype4	18	4	0

Figure S48. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'RNAseq cHierClus subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 1 (Fisher's exact test)

Table S54. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients	NO	YES
ALL	82	1
subtype1	17	0
subtype2	16	0
subtype3	27	1
subtype4	22	0

Figure S49. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'RNAseq cHierClus subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 1 (Fisher's exact test)

Table S55. Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

nPatients	NO	YES
ALL	82	1
subtype1	17	0
subtype2	16	0
subtype3	27	1
subtype4	22	0

Figure S50. Get High-res Image Clustering Approach #5: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

Clustering Approach #6: 'MIRseq CNMF subtypes'

Table S56. Get Full Table Description of clustering approach #6: 'MIRseq CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	54	114	87

'MIRseq CNMF subtypes' versus 'Time to Death'

P value = 0.0895 (logrank test)

Table S57. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	141	20	0.9 - 72.1 (9.1)
subtype1	32	8	0.9 - 56.4 (11.8)
subtype2	76	10	0.9 - 72.1 (14.1)
subtype3	33	2	0.9 - 38.9 (1.0)

Figure S51. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq CNMF subtypes' versus 'AGE'

P value = 0.82 (ANOVA)

Table S58. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	255	68.6 (12.2)
subtype1	54	68.1 (11.0)
subtype2	114	68.4 (12.6)
subtype3	87	69.3 (12.4)

Figure S52. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

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

P value = 0.965 (Fisher's exact test)

Table S59. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	COLON	RECTUM
ALL	186	67
subtype1	40	13
subtype2	83	31
subtype3	63	23

Figure S53. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'MIRseq CNMF subtypes' versus 'GENDER'

P value = 0.653 (Fisher's exact test)

Table S60. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	128	127
subtype1	24	30
subtype2	59	55
subtype3	45	42

Figure S54. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #4: 'GENDER'

'MIRseq CNMF subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 0.942 (Chi-square test)

Table S61. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients	COLON ADENOCARCINOMA	COLON MUCINOUS ADENOCARCINOMA	RECTAL ADENOCARCINOMA	RECTAL MUCINOUS ADENOCARCINOMA
ALL	158	26	56	8
subtype1	36	4	11	1
subtype2	68	14	25	4
subtype3	54	8	20	3

Figure S55. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'MIRseq CNMF subtypes' versus 'PATHOLOGY.T'

P value = 0.331 (Chi-square test)

Table S62. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients	T0+T1	T2	T3	T4
ALL	11	52	169	21
subtype1	2	9	34	9
subtype2	5	24	76	8
subtype3	4	19	59	4

Figure S56. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

'MIRseq CNMF subtypes' versus 'PATHOLOGY.N'

P value = 0.0292 (Chi-square test)

Table S63. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients	N0	N1	N2
ALL	152	55	47
subtype1	24	20	9
subtype2	71	20	23
subtype3	57	15	15

Figure S57. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

'MIRseq CNMF subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 0.1 (Chi-square test)

Table S64. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients	M0	M1	M1A
ALL	208	41	2
subtype1	37	14	1
subtype2	99	14	0
subtype3	72	13	1

Figure S58. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'MIRseq CNMF subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.553 (Fisher's exact test)

Table S65. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients	NO	YES
ALL	254	1
subtype1	54	0
subtype2	114	0
subtype3	86	1

Figure S59. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'MIRseq CNMF subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 0.0117 (Fisher's exact test)

Table S66. Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

nPatients	NO	YES
ALL	241	14
subtype1	47	7
subtype2	108	6
subtype3	86	1

Figure S60. Get High-res Image Clustering Approach #6: 'MIRseq CNMF subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

Clustering Approach #7: 'MIRseq cHierClus subtypes'

Table S67. Get Full Table Description of clustering approach #7: 'MIRseq cHierClus subtypes'

Cluster Labels	1	2
Number of samples	80	175

'MIRseq cHierClus subtypes' versus 'Time to Death'

P value = 0.738 (logrank test)

Table S68. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	141	20	0.9 - 72.1 (9.1)
subtype1	36	2	0.9 - 35.0 (1.0)
subtype2	105	18	0.9 - 72.1 (13.0)

Figure S61. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq cHierClus subtypes' versus 'AGE'

P value = 0.472 (t-test)

Table S69. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

	nPatients	Mean (Std.Dev)
ALL	255	68.6 (12.2)
subtype1	80	67.8 (12.7)
subtype2	175	69.0 (12.0)

Figure S62. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

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

P value = 0.76 (Fisher's exact test)

Table S70. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients	COLON	RECTUM
ALL	186	67
subtype1	57	22
subtype2	129	45

Figure S63. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'MIRseq cHierClus subtypes' versus 'GENDER'

P value = 0.42 (Fisher's exact test)

Table S71. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

nPatients	FEMALE	MALE
ALL	128	127
subtype1	37	43
subtype2	91	84

Figure S64. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

'MIRseq cHierClus subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 0.513 (Chi-square test)

Table S72. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients	COLON ADENOCARCINOMA	COLON MUCINOUS ADENOCARCINOMA	RECTAL ADENOCARCINOMA	RECTAL MUCINOUS ADENOCARCINOMA
ALL	158	26	56	8
subtype1	50	6	19	1
subtype2	108	20	37	7

Figure S65. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'MIRseq cHierClus subtypes' versus 'PATHOLOGY.T'

P value = 0.208 (Chi-square test)

Table S73. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients	T0+T1	T2	T3	T4
ALL	11	52	169	21
subtype1	5	19	52	3
subtype2	6	33	117	18

Figure S66. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

'MIRseq cHierClus subtypes' versus 'PATHOLOGY.N'

P value = 0.946 (Chi-square test)

Table S74. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients	N0	N1	N2
ALL	152	55	47
subtype1	49	17	14
subtype2	103	38	33

Figure S67. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

'MIRseq cHierClus subtypes' versus 'PATHOLOGICSPREAD(M)'

P value = 0.313 (Chi-square test)

Table S75. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients	M0	M1	M1A
ALL	208	41	2
subtype1	69	9	1
subtype2	139	32	1

Figure S68. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'MIRseq cHierClus subtypes' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.314 (Fisher's exact test)

Table S76. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients	NO	YES
ALL	254	1
subtype1	79	1
subtype2	175	0

Figure S69. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #9: 'RADIATIONS.RADIATION.REGIMENINDICATION'

'MIRseq cHierClus subtypes' versus 'NEOADJUVANT.THERAPY'

P value = 0.237 (Fisher's exact test)

Table S77. Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

nPatients	NO	YES
ALL	241	14
subtype1	78	2
subtype2	163	12

Figure S70. Get High-res Image Clustering Approach #7: 'MIRseq cHierClus subtypes' versus Clinical Feature #10: 'NEOADJUVANT.THERAPY'

Methods & Data

Input

Cluster data file = COADREAD.mergedcluster.txt
Clinical data file = COADREAD.clin.merged.picked.txt
Number of patients = 585
Number of clustering approaches = 7
Number of selected clinical features = 10
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

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

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)

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

Made with Nozzle