Colon/Rectal Adenocarcinoma: 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 10 different clustering approaches and 10 clinical features across 588 patients, 37 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',  'PATHOLOGICSPREAD(M)', and 'TUMOR.STAGE'.

  • Consensus hierarchical clustering analysis on array-based mRNA expression data identified 3 subtypes that correlate to 'PRIMARY.SITE.OF.DISEASE' and 'HISTOLOGICAL.TYPE'.

  • 4 subtypes identified in current cancer cohort by 'CN CNMF'. These subtypes correlate to 'PRIMARY.SITE.OF.DISEASE',  'HISTOLOGICAL.TYPE',  'PATHOLOGY.N', and 'TUMOR.STAGE'.

  • 3 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. These subtypes correlate to 'AGE',  'PRIMARY.SITE.OF.DISEASE',  'GENDER', and 'HISTOLOGICAL.TYPE'.

  • CNMF clustering analysis on RPPA data identified 4 subtypes that correlate to 'PATHOLOGY.N',  'PATHOLOGICSPREAD(M)', and 'TUMOR.STAGE'.

  • Consensus hierarchical clustering analysis on RPPA data identified 4 subtypes that correlate to 'PATHOLOGY.N',  'PATHOLOGICSPREAD(M)', and 'TUMOR.STAGE'.

  • CNMF clustering analysis on sequencing-based mRNA expression data identified 4 subtypes that correlate to 'PRIMARY.SITE.OF.DISEASE',  'GENDER',  'HISTOLOGICAL.TYPE',  'PATHOLOGICSPREAD(M)', and 'TUMOR.STAGE'.

  • Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 3 subtypes that correlate to 'AGE',  'PRIMARY.SITE.OF.DISEASE', and 'HISTOLOGICAL.TYPE'.

  • CNMF clustering analysis on sequencing-based miR expression data identified 3 subtypes that correlate to 'AGE',  'HISTOLOGICAL.TYPE',  'PATHOLOGY.N',  'PATHOLOGICSPREAD(M)', and 'TUMOR.STAGE'.

  • Consensus hierarchical clustering analysis on sequencing-based miR expression data identified 3 subtypes that correlate to 'AGE',  'PATHOLOGY.T', and 'PATHOLOGICSPREAD(M)'.

Results
Overview of the results

Table 1.  Get Full Table Overview of the association between subtypes identified by 10 different clustering approaches and 10 clinical features. Shown in the table are P values from statistical tests. Thresholded by P value < 0.05, 37 significant findings detected.

Clinical
Features 		Time
to
Death 		AGE PRIMARY
SITE
OF
DISEASE 		GENDER HISTOLOGICAL
TYPE 		PATHOLOGY
T 		PATHOLOGY
N 		PATHOLOGICSPREAD(M) TUMOR
STAGE 		RADIATIONS
RADIATION
REGIMENINDICATION
Statistical Tests logrank test ANOVA Fisher's exact test Fisher's exact test Chi-square test Chi-square test Chi-square test Chi-square test Chi-square test Fisher's exact test
mRNA CNMF subtypes 0.546 0.596 0.0126 0.072 6.39e-07 0.184 0.00595 0.00173 0.0119 0.393
mRNA cHierClus subtypes 0.97 0.148 0.035 0.796 3.6e-08 0.866 0.65 0.498 0.633 0.496
CN CNMF 0.85 0.628 7.44e-05 0.804 4.28e-13 0.934 0.00155 0.2 0.00415 0.216
METHLYATION CNMF 0.964 0.00171 0.00302 0.0473 0.00521 0.466 0.497 0.748 0.39 0.0619
RPPA CNMF subtypes 0.752 0.467 0.469 0.474 0.447 0.455 0.0253 0.0189 0.0429 0.826
RPPA cHierClus subtypes 0.424 0.505 0.258 0.121 0.272 0.527 0.0289 0.00181 0.0126 0.165
RNAseq CNMF subtypes 0.958 0.441 0.00393 0.0254 1.52e-08 0.28 0.0659 0.0049 0.0118 0.0712
RNAseq cHierClus subtypes 0.839 0.0148 0.000199 0.527 1.17e-14 0.349 0.324 0.285 0.0536 0.643
MIRseq CNMF subtypes 0.525 0.00518 0.0828 0.255 0.000466 0.0523 0.019 3.15e-08 0.0489 0.0599
MIRseq cHierClus subtypes 0.766 0.00244 0.81 0.417 0.382 0.0053 0.236 1.91e-07 0.0748 0.365
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 = 0.546 (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 107 117
subtype1 33 30
subtype2 29 19
subtype3 31 42
subtype4 14 26

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 'TUMOR.STAGE'

P value = 0.0119 (Chi-square test)

Table S10.  Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 46 86 55 32
subtype1 9 16 19 17
subtype2 8 21 12 5
subtype3 19 28 18 8
subtype4 10 21 6 2

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

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

P value = 0.393 (Fisher's exact test)

Table S11.  Clustering Approach #1: 'mRNA CNMF subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

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

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

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 = 0.97 (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 107 117
subtype1 22 25
subtype2 33 31
subtype3 52 61

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 'TUMOR.STAGE'

P value = 0.633 (Chi-square test)

Table S21.  Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 46 86 55 32
subtype1 9 19 11 8
subtype2 13 29 15 5
subtype3 24 38 29 19

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

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

P value = 0.496 (Fisher's exact test)

Table S22.  Clustering Approach #2: 'mRNA cHierClus subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

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

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

Clustering Approach #3: 'CN CNMF'

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

Cluster Labels 1 2 3 4
Number of samples 200 234 100 41
'CN CNMF' versus 'Time to Death'

P value = 0.85 (logrank test)

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

nPatients nDeath Duration Range (Median), Month
ALL 444 60 0.1 - 135.5 (7.0)
subtype1 150 22 0.1 - 129.1 (8.0)
subtype2 191 22 0.1 - 135.5 (6.0)
subtype3 73 12 0.2 - 112.7 (6.6)
subtype4 30 4 0.3 - 87.8 (13.3)

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

'CN CNMF' versus 'AGE'

P value = 0.628 (ANOVA)

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

nPatients Mean (Std.Dev)
ALL 574 66.6 (12.7)
subtype1 200 66.8 (11.2)
subtype2 233 66.8 (14.1)
subtype3 100 66.8 (12.1)
subtype4 41 64.1 (13.4)

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

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

P value = 7.44e-05 (Fisher's exact test)

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

nPatients COLON RECTUM
ALL 412 159
subtype1 124 72
subtype2 192 42
subtype3 66 34
subtype4 30 11

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

'CN CNMF' versus 'GENDER'

P value = 0.804 (Fisher's exact test)

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

nPatients FEMALE MALE
ALL 267 308
subtype1 88 112
subtype2 113 121
subtype3 48 52
subtype4 18 23

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

'CN CNMF' versus 'HISTOLOGICAL.TYPE'

P value = 4.28e-13 (Chi-square test)

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

nPatients COLON ADENOCARCINOMA COLON MUCINOUS ADENOCARCINOMA RECTAL ADENOCARCINOMA RECTAL MUCINOUS ADENOCARCINOMA
ALL 356 54 143 13
subtype1 121 3 72 0
subtype2 144 46 32 9
subtype3 61 5 28 4
subtype4 30 0 11 0

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

'CN CNMF' versus 'PATHOLOGY.T'

P value = 0.934 (Chi-square test)

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

nPatients T0+T1 T2 T3 T4
ALL 20 101 392 59
subtype1 7 31 140 22
subtype2 10 44 154 24
subtype3 2 18 69 11
subtype4 1 8 29 2

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

'CN CNMF' versus 'PATHOLOGY.N'

P value = 0.00155 (Chi-square test)

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

nPatients N0 N1 N2
ALL 328 140 103
subtype1 96 64 38
subtype2 157 45 32
subtype3 51 22 26
subtype4 24 9 7

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

'CN CNMF' versus 'PATHOLOGICSPREAD(M)'

P value = 0.2 (Chi-square test)

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

nPatients M0 M1 M1A M1B MX
ALL 435 71 9 1 49
subtype1 143 35 4 0 16
subtype2 189 16 2 1 21
subtype3 72 15 2 0 10
subtype4 31 5 1 0 2

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

'CN CNMF' versus 'TUMOR.STAGE'

P value = 0.00415 (Chi-square test)

Table S32.  Clustering Approach #3: 'CN CNMF' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 97 207 166 81
subtype1 28 58 68 38
subtype2 47 100 58 19
subtype3 16 32 32 18
subtype4 6 17 8 6

Figure S29.  Get High-res Image Clustering Approach #3: 'CN CNMF' versus Clinical Feature #9: 'TUMOR.STAGE'

'CN CNMF' versus 'RADIATIONS.RADIATION.REGIMENINDICATION'

P value = 0.216 (Fisher's exact test)

Table S33.  Clustering Approach #3: 'CN CNMF' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 9 566
subtype1 3 197
subtype2 2 232
subtype3 2 98
subtype4 2 39

Figure S30.  Get High-res Image Clustering Approach #3: 'CN CNMF' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

Clustering Approach #4: 'METHLYATION CNMF'

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

Cluster Labels 1 2 3
Number of samples 146 113 91
'METHLYATION CNMF' versus 'Time to Death'

P value = 0.964 (logrank test)

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

nPatients nDeath Duration Range (Median), Month
ALL 325 44 0.1 - 135.5 (7.0)
subtype1 136 20 0.1 - 135.5 (7.5)
subtype2 104 16 0.1 - 129.1 (7.5)
subtype3 85 8 0.1 - 102.4 (5.9)

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

'METHLYATION CNMF' versus 'AGE'

P value = 0.00171 (ANOVA)

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

nPatients Mean (Std.Dev)
ALL 349 64.9 (13.0)
subtype1 146 65.2 (12.6)
subtype2 113 67.4 (13.0)
subtype3 90 61.0 (12.9)

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

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

P value = 0.00302 (Fisher's exact test)

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

nPatients COLON RECTUM
ALL 255 93
subtype1 95 50
subtype2 95 18
subtype3 65 25

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

'METHLYATION CNMF' versus 'GENDER'

P value = 0.0473 (Fisher's exact test)

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

nPatients FEMALE MALE
ALL 157 193
subtype1 62 84
subtype2 61 52
subtype3 34 57

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

'METHLYATION CNMF' versus 'HISTOLOGICAL.TYPE'

P value = 0.00521 (Chi-square test)

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

nPatients COLON ADENOCARCINOMA COLON MUCINOUS ADENOCARCINOMA RECTAL ADENOCARCINOMA RECTAL MUCINOUS ADENOCARCINOMA
ALL 225 30 87 6
subtype1 89 6 48 2
subtype2 82 13 15 3
subtype3 54 11 24 1

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

'METHLYATION CNMF' versus 'PATHOLOGY.T'

P value = 0.466 (Chi-square test)

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

nPatients T0+T1 T2 T3 T4
ALL 11 51 244 42
subtype1 4 23 101 17
subtype2 4 20 78 10
subtype3 3 8 65 15

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

'METHLYATION CNMF' versus 'PATHOLOGY.N'

P value = 0.497 (Chi-square test)

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

nPatients N0 N1 N2
ALL 193 94 58
subtype1 75 43 23
subtype2 70 24 19
subtype3 48 27 16

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

'METHLYATION CNMF' versus 'PATHOLOGICSPREAD(M)'

P value = 0.748 (Chi-square test)

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

nPatients M0 M1 M1A M1B MX
ALL 247 37 8 1 50
subtype1 99 18 5 0 22
subtype2 82 10 1 1 16
subtype3 66 9 2 0 12

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

'METHLYATION CNMF' versus 'TUMOR.STAGE'

P value = 0.39 (Chi-square test)

Table S43.  Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 48 126 107 49
subtype1 18 48 49 23
subtype2 21 45 28 14
subtype3 9 33 30 12

Figure S39.  Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #9: 'TUMOR.STAGE'

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

P value = 0.0619 (Fisher's exact test)

Table S44.  Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 8 342
subtype1 4 142
subtype2 0 113
subtype3 4 87

Figure S40.  Get High-res Image Clustering Approach #4: 'METHLYATION CNMF' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

Clustering Approach #5: 'RPPA CNMF subtypes'

Table S45.  Get Full Table Description of clustering approach #5: 'RPPA CNMF subtypes'

Cluster Labels 1 2 3 4
Number of samples 96 160 20 123
'RPPA CNMF subtypes' versus 'Time to Death'

P value = 0.752 (logrank test)

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

nPatients nDeath Duration Range (Median), Month
ALL 308 30 0.1 - 121.1 (6.0)
subtype1 70 4 0.1 - 87.8 (4.4)
subtype2 117 13 0.1 - 121.1 (6.0)
subtype3 13 0 0.2 - 41.0 (7.6)
subtype4 108 13 0.1 - 105.3 (7.0)

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

'RPPA CNMF subtypes' versus 'AGE'

P value = 0.467 (ANOVA)

Table S47.  Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 398 66.2 (12.8)
subtype1 96 66.6 (12.0)
subtype2 159 66.1 (12.9)
subtype3 20 70.3 (10.0)
subtype4 123 65.5 (13.8)

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

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

P value = 0.469 (Fisher's exact test)

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

nPatients COLON RECTUM
ALL 269 130
subtype1 64 32
subtype2 106 54
subtype3 11 9
subtype4 88 35

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

'RPPA CNMF subtypes' versus 'GENDER'

P value = 0.474 (Fisher's exact test)

Table S49.  Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'GENDER'

nPatients FEMALE MALE
ALL 187 212
subtype1 39 57
subtype2 81 79
subtype3 10 10
subtype4 57 66

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

'RPPA CNMF subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 0.447 (Chi-square test)

Table S50.  Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients COLON ADENOCARCINOMA COLON MUCINOUS ADENOCARCINOMA RECTAL ADENOCARCINOMA RECTAL MUCINOUS ADENOCARCINOMA
ALL 238 30 117 10
subtype1 54 10 26 4
subtype2 99 7 49 4
subtype3 9 2 8 1
subtype4 76 11 34 1

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

'RPPA CNMF subtypes' versus 'PATHOLOGY.T'

P value = 0.455 (Chi-square test)

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

nPatients T0+T1 T2 T3 T4
ALL 10 68 276 41
subtype1 0 14 68 13
subtype2 5 26 110 16
subtype3 0 4 16 0
subtype4 5 24 82 12

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

'RPPA CNMF subtypes' versus 'PATHOLOGY.N'

P value = 0.0253 (Chi-square test)

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

nPatients N0 N1 N2
ALL 224 103 68
subtype1 45 29 20
subtype2 82 47 31
subtype3 15 2 2
subtype4 82 25 15

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

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

P value = 0.0189 (Chi-square test)

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

nPatients M0 M1 M1A M1B MX
ALL 300 49 8 1 36
subtype1 72 12 5 0 5
subtype2 111 29 0 1 17
subtype3 18 1 0 0 1
subtype4 99 7 3 0 13

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

'RPPA CNMF subtypes' versus 'TUMOR.STAGE'

P value = 0.0429 (Chi-square test)

Table S54.  Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 61 147 119 57
subtype1 11 31 32 16
subtype2 23 50 51 30
subtype3 4 11 3 1
subtype4 23 55 33 10

Figure S49.  Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

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

P value = 0.826 (Fisher's exact test)

Table S55.  Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 8 391
subtype1 3 93
subtype2 3 157
subtype3 0 20
subtype4 2 121

Figure S50.  Get High-res Image Clustering Approach #5: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

Clustering Approach #6: 'RPPA cHierClus subtypes'

Table S56.  Get Full Table Description of clustering approach #6: 'RPPA cHierClus subtypes'

Cluster Labels 1 2 3 4
Number of samples 210 34 72 83
'RPPA cHierClus subtypes' versus 'Time to Death'

P value = 0.424 (logrank test)

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

nPatients nDeath Duration Range (Median), Month
ALL 308 30 0.1 - 121.1 (6.0)
subtype1 148 15 0.1 - 121.1 (5.9)
subtype2 22 3 0.1 - 105.3 (8.1)
subtype3 65 2 0.1 - 87.8 (5.4)
subtype4 73 10 0.1 - 100.0 (6.1)

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

'RPPA cHierClus subtypes' versus 'AGE'

P value = 0.505 (ANOVA)

Table S58.  Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 398 66.2 (12.8)
subtype1 209 66.8 (13.2)
subtype2 34 67.9 (11.8)
subtype3 72 65.1 (12.1)
subtype4 83 65.1 (12.8)

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

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

P value = 0.258 (Fisher's exact test)

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

nPatients COLON RECTUM
ALL 269 130
subtype1 138 72
subtype2 28 6
subtype3 49 23
subtype4 54 29

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

'RPPA cHierClus subtypes' versus 'GENDER'

P value = 0.121 (Fisher's exact test)

Table S60.  Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

nPatients FEMALE MALE
ALL 187 212
subtype1 105 105
subtype2 19 15
subtype3 33 39
subtype4 30 53

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

'RPPA cHierClus subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 0.272 (Chi-square test)

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

nPatients COLON ADENOCARCINOMA COLON MUCINOUS ADENOCARCINOMA RECTAL ADENOCARCINOMA RECTAL MUCINOUS ADENOCARCINOMA
ALL 238 30 117 10
subtype1 125 12 63 6
subtype2 25 3 5 1
subtype3 40 9 20 3
subtype4 48 6 29 0

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

'RPPA cHierClus subtypes' versus 'PATHOLOGY.T'

P value = 0.527 (Chi-square test)

Table S62.  Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients T0+T1 T2 T3 T4
ALL 10 68 276 41
subtype1 5 35 145 22
subtype2 1 3 28 2
subtype3 1 11 48 11
subtype4 3 19 55 6

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

'RPPA cHierClus subtypes' versus 'PATHOLOGY.N'

P value = 0.0289 (Chi-square test)

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

nPatients N0 N1 N2
ALL 224 103 68
subtype1 107 62 40
subtype2 22 7 5
subtype3 35 20 15
subtype4 60 14 8

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

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

P value = 0.00181 (Chi-square test)

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

nPatients M0 M1 M1A M1B MX
ALL 300 49 8 1 36
subtype1 153 37 0 1 19
subtype2 29 0 0 0 3
subtype3 53 8 5 0 4
subtype4 65 4 3 0 10

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

'RPPA cHierClus subtypes' versus 'TUMOR.STAGE'

P value = 0.0126 (Chi-square test)

Table S65.  Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 61 147 119 57
subtype1 31 69 65 37
subtype2 3 18 12 0
subtype3 9 22 24 13
subtype4 18 38 18 7

Figure S59.  Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

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

P value = 0.165 (Fisher's exact test)

Table S66.  Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 8 391
subtype1 3 207
subtype2 0 34
subtype3 4 68
subtype4 1 82

Figure S60.  Get High-res Image Clustering Approach #6: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

Clustering Approach #7: 'RNAseq CNMF subtypes'

Table S67.  Get Full Table Description of clustering approach #7: 'RNAseq CNMF subtypes'

Cluster Labels 1 2 3 4
Number of samples 96 60 70 38
'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.958 (logrank test)

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

nPatients nDeath Duration Range (Median), Month
ALL 151 20 0.9 - 72.1 (12.0)
subtype1 55 9 0.9 - 72.1 (13.9)
subtype2 39 4 1.0 - 64.0 (10.9)
subtype3 42 6 0.9 - 61.9 (12.0)
subtype4 15 1 0.9 - 34.0 (1.0)

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

'RNAseq CNMF subtypes' versus 'AGE'

P value = 0.441 (ANOVA)

Table S69.  Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 264 68.9 (11.9)
subtype1 96 67.5 (11.5)
subtype2 60 70.2 (12.7)
subtype3 70 70.0 (12.2)
subtype4 38 68.1 (11.3)

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

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

P value = 0.00393 (Fisher's exact test)

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

nPatients COLON RECTUM
ALL 191 71
subtype1 60 35
subtype2 53 7
subtype3 53 17
subtype4 25 12

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

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 0.0254 (Fisher's exact test)

Table S71.  Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'GENDER'

nPatients FEMALE MALE
ALL 132 132
subtype1 51 45
subtype2 38 22
subtype3 29 41
subtype4 14 24

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

'RNAseq CNMF subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 1.52e-08 (Chi-square test)

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

nPatients COLON ADENOCARCINOMA COLON MUCINOUS ADENOCARCINOMA RECTAL ADENOCARCINOMA RECTAL MUCINOUS ADENOCARCINOMA
ALL 162 27 60 8
subtype1 60 0 34 0
subtype2 35 17 4 2
subtype3 43 9 11 5
subtype4 24 1 11 1

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

'RNAseq CNMF subtypes' versus 'PATHOLOGY.T'

P value = 0.28 (Chi-square test)

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

nPatients T0+T1 T2 T3 T4
ALL 11 53 176 22
subtype1 2 22 61 11
subtype2 1 9 42 6
subtype3 5 13 47 5
subtype4 3 9 26 0

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

'RNAseq CNMF subtypes' versus 'PATHOLOGY.N'

P value = 0.0659 (Chi-square test)

Table S74.  Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients N0 N1 N2
ALL 158 57 48
subtype1 49 27 20
subtype2 34 10 15
subtype3 50 14 6
subtype4 25 6 7

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

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

P value = 0.0049 (Chi-square test)

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

nPatients M0 M1 M1A
ALL 220 39 2
subtype1 70 24 1
subtype2 50 10 0
subtype3 66 2 1
subtype4 34 3 0

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

'RNAseq CNMF subtypes' versus 'TUMOR.STAGE'

P value = 0.0118 (Chi-square test)

Table S76.  Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 53 98 68 37
subtype1 16 28 26 23
subtype2 8 25 16 8
subtype3 17 32 16 3
subtype4 12 13 10 3

Figure S69.  Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

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

P value = 0.0712 (Fisher's exact test)

Table S77.  Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 2 262
subtype1 0 96
subtype2 2 58
subtype3 0 70
subtype4 0 38

Figure S70.  Get High-res Image Clustering Approach #7: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

Clustering Approach #8: 'RNAseq cHierClus subtypes'

Table S78.  Get Full Table Description of clustering approach #8: 'RNAseq cHierClus subtypes'

Cluster Labels 1 2 3
Number of samples 62 158 44
'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.839 (logrank test)

Table S79.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 151 20 0.9 - 72.1 (12.0)
subtype1 37 5 1.0 - 49.0 (10.3)
subtype2 86 11 0.9 - 72.1 (12.0)
subtype3 28 4 0.9 - 61.9 (11.5)

Figure S71.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'AGE'

P value = 0.0148 (ANOVA)

Table S80.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 264 68.9 (11.9)
subtype1 62 72.7 (12.5)
subtype2 158 67.6 (11.6)
subtype3 44 68.1 (11.5)

Figure S72.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

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

P value = 0.000199 (Fisher's exact test)

Table S81.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients COLON RECTUM
ALL 191 71
subtype1 57 5
subtype2 104 52
subtype3 30 14

Figure S73.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 0.527 (Fisher's exact test)

Table S82.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

nPatients FEMALE MALE
ALL 132 132
subtype1 35 27
subtype2 76 82
subtype3 21 23

Figure S74.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

'RNAseq cHierClus subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 1.17e-14 (Chi-square test)

Table S83.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients COLON ADENOCARCINOMA COLON MUCINOUS ADENOCARCINOMA RECTAL ADENOCARCINOMA RECTAL MUCINOUS ADENOCARCINOMA
ALL 162 27 60 8
subtype1 37 19 3 1
subtype2 103 1 50 1
subtype3 22 7 7 6

Figure S75.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY.T'

P value = 0.349 (Chi-square test)

Table S84.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients T0+T1 T2 T3 T4
ALL 11 53 176 22
subtype1 1 9 42 8
subtype2 7 35 103 13
subtype3 3 9 31 1

Figure S76.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY.N'

P value = 0.324 (Chi-square test)

Table S85.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients N0 N1 N2
ALL 158 57 48
subtype1 39 12 11
subtype2 87 38 32
subtype3 32 7 5

Figure S77.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

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

P value = 0.285 (Chi-square test)

Table S86.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients M0 M1 M1A
ALL 220 39 2
subtype1 54 6 1
subtype2 126 29 1
subtype3 40 4 0

Figure S78.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'RNAseq cHierClus subtypes' versus 'TUMOR.STAGE'

P value = 0.0536 (Chi-square test)

Table S87.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 53 98 68 37
subtype1 8 30 16 6
subtype2 33 48 45 27
subtype3 12 20 7 4

Figure S79.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

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

P value = 0.643 (Fisher's exact test)

Table S88.  Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 2 262
subtype1 1 61
subtype2 1 157
subtype3 0 44

Figure S80.  Get High-res Image Clustering Approach #8: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

Clustering Approach #9: 'MIRseq CNMF subtypes'

Table S89.  Get Full Table Description of clustering approach #9: 'MIRseq CNMF subtypes'

Cluster Labels 1 2 3
Number of samples 245 222 83
'MIRseq CNMF subtypes' versus 'Time to Death'

P value = 0.525 (logrank test)

Table S90.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 415 59 0.1 - 135.5 (7.0)
subtype1 132 18 0.2 - 72.1 (7.3)
subtype2 206 32 0.1 - 135.5 (7.8)
subtype3 77 9 0.1 - 100.0 (4.8)

Figure S81.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq CNMF subtypes' versus 'AGE'

P value = 0.00518 (ANOVA)

Table S91.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 549 66.8 (12.7)
subtype1 245 68.7 (12.1)
subtype2 221 65.4 (12.6)
subtype3 83 64.8 (14.1)

Figure S82.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #2: 'AGE'

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

P value = 0.0828 (Fisher's exact test)

Table S92.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients COLON RECTUM
ALL 406 140
subtype1 177 66
subtype2 160 61
subtype3 69 13

Figure S83.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'MIRseq CNMF subtypes' versus 'GENDER'

P value = 0.255 (Fisher's exact test)

Table S93.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #4: 'GENDER'

nPatients FEMALE MALE
ALL 258 292
subtype1 120 125
subtype2 106 116
subtype3 32 51

Figure S84.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #4: 'GENDER'

'MIRseq CNMF subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 0.000466 (Chi-square test)

Table S94.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients COLON ADENOCARCINOMA COLON MUCINOUS ADENOCARCINOMA RECTAL ADENOCARCINOMA RECTAL MUCINOUS ADENOCARCINOMA
ALL 349 55 127 10
subtype1 151 24 56 7
subtype2 147 13 60 1
subtype3 51 18 11 2

Figure S85.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'MIRseq CNMF subtypes' versus 'PATHOLOGY.T'

P value = 0.0523 (Chi-square test)

Table S95.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients T0+T1 T2 T3 T4
ALL 20 95 376 55
subtype1 10 50 162 21
subtype2 9 37 155 19
subtype3 1 8 59 15

Figure S86.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

'MIRseq CNMF subtypes' versus 'PATHOLOGY.N'

P value = 0.019 (Chi-square test)

Table S96.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients N0 N1 N2
ALL 315 132 98
subtype1 150 48 47
subtype2 128 61 30
subtype3 37 23 21

Figure S87.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

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

P value = 3.15e-08 (Chi-square test)

Table S97.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients M0 M1 M1A M1B MX
ALL 412 68 9 1 50
subtype1 199 40 1 0 1
subtype2 157 19 6 1 35
subtype3 56 9 2 0 14

Figure S88.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'MIRseq CNMF subtypes' versus 'TUMOR.STAGE'

P value = 0.0489 (Chi-square test)

Table S98.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 93 199 156 77
subtype1 50 91 58 37
subtype2 38 81 68 28
subtype3 5 27 30 12

Figure S89.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

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

P value = 0.0599 (Fisher's exact test)

Table S99.  Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 9 541
subtype1 2 243
subtype2 3 219
subtype3 4 79

Figure S90.  Get High-res Image Clustering Approach #9: 'MIRseq CNMF subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

Clustering Approach #10: 'MIRseq cHierClus subtypes'

Table S100.  Get Full Table Description of clustering approach #10: 'MIRseq cHierClus subtypes'

Cluster Labels 1 2 3
Number of samples 236 33 281
'MIRseq cHierClus subtypes' versus 'Time to Death'

P value = 0.766 (logrank test)

Table S101.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 415 59 0.1 - 135.5 (7.0)
subtype1 125 17 0.6 - 72.1 (9.4)
subtype2 30 5 0.5 - 112.7 (9.3)
subtype3 260 37 0.1 - 135.5 (6.9)

Figure S91.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq cHierClus subtypes' versus 'AGE'

P value = 0.00244 (ANOVA)

Table S102.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

nPatients Mean (Std.Dev)
ALL 549 66.8 (12.7)
subtype1 236 68.9 (12.2)
subtype2 33 67.2 (11.3)
subtype3 280 65.0 (13.1)

Figure S92.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #2: 'AGE'

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

P value = 0.81 (Fisher's exact test)

Table S103.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

nPatients COLON RECTUM
ALL 406 140
subtype1 177 57
subtype2 25 8
subtype3 204 75

Figure S93.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #3: 'PRIMARY.SITE.OF.DISEASE'

'MIRseq cHierClus subtypes' versus 'GENDER'

P value = 0.417 (Fisher's exact test)

Table S104.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

nPatients FEMALE MALE
ALL 258 292
subtype1 115 121
subtype2 18 15
subtype3 125 156

Figure S94.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #4: 'GENDER'

'MIRseq cHierClus subtypes' versus 'HISTOLOGICAL.TYPE'

P value = 0.382 (Chi-square test)

Table S105.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

nPatients COLON ADENOCARCINOMA COLON MUCINOUS ADENOCARCINOMA RECTAL ADENOCARCINOMA RECTAL MUCINOUS ADENOCARCINOMA
ALL 349 55 127 10
subtype1 149 26 47 7
subtype2 22 3 7 1
subtype3 178 26 73 2

Figure S95.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #5: 'HISTOLOGICAL.TYPE'

'MIRseq cHierClus subtypes' versus 'PATHOLOGY.T'

P value = 0.0053 (Chi-square test)

Table S106.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

nPatients T0+T1 T2 T3 T4
ALL 20 95 376 55
subtype1 9 49 156 20
subtype2 5 4 21 3
subtype3 6 42 199 32

Figure S96.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY.T'

'MIRseq cHierClus subtypes' versus 'PATHOLOGY.N'

P value = 0.236 (Chi-square test)

Table S107.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

nPatients N0 N1 N2
ALL 315 132 98
subtype1 140 51 44
subtype2 24 5 4
subtype3 151 76 50

Figure S97.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #7: 'PATHOLOGY.N'

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

P value = 1.91e-07 (Chi-square test)

Table S108.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

nPatients M0 M1 M1A M1B MX
ALL 412 68 9 1 50
subtype1 191 38 2 0 1
subtype2 25 4 0 0 3
subtype3 196 26 7 1 46

Figure S98.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #8: 'PATHOLOGICSPREAD(M)'

'MIRseq cHierClus subtypes' versus 'TUMOR.STAGE'

P value = 0.0748 (Chi-square test)

Table S109.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

nPatients I II III IV
ALL 93 199 156 77
subtype1 48 83 60 35
subtype2 9 14 5 4
subtype3 36 102 91 38

Figure S99.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #9: 'TUMOR.STAGE'

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

P value = 0.365 (Fisher's exact test)

Table S110.  Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

nPatients NO YES
ALL 9 541
subtype1 2 234
subtype2 0 33
subtype3 7 274

Figure S100.  Get High-res Image Clustering Approach #10: 'MIRseq cHierClus subtypes' versus Clinical Feature #10: 'RADIATIONS.RADIATION.REGIMENINDICATION'

Methods & Data
Input

  • Cluster data file = COADREAD-TP.mergedcluster.txt

  • Clinical data file = COADREAD-TP.clin.merged.picked.txt

  • Number of patients = 588

  • Number of clustering approaches = 10

  • 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

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)
[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)
Copyright © 2013 Broad Institute TCGA GDAC as part of the TCGA Research Network. All rights reserved.
Made with Nozzle