Correlation between gene mutation status and molecular subtypes
Overview
Introduction

This pipeline computes the correlation between significantly recurrent gene mutations and molecular subtypes.

Summary

Testing the association between mutation status of 21 genes and 8 molecular subtypes across 224 patients, 9 significant findings detected with P value < 0.05 and Q value < 0.25.

  • FBXW7 mutation correlated to 'CN_CNMF'.

  • BRAF mutation correlated to 'MRNA_CNMF',  'MRNA_CHIERARCHICAL', and 'CN_CNMF'.

  • KRAS mutation correlated to 'MRNA_CNMF' and 'MRNA_CHIERARCHICAL'.

  • TP53 mutation correlated to 'MRNA_CHIERARCHICAL' and 'CN_CNMF'.

  • PIK3CA mutation correlated to 'CN_CNMF'.

Results
Overview of the results

Table 1.  Get Full Table Overview of the association between mutation status of 21 genes and 8 molecular subtypes. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, 9 significant findings detected.

Clinical
Features 		MRNA
CNMF 		MRNA
CHIERARCHICAL 		CN
CNMF 		METHLYATION
CNMF 		RPPA
CNMF 		RPPA
CHIERARCHICAL 		MIRSEQ
CNMF 		MIRSEQ
CHIERARCHICAL
nMutated (%) nWild-Type Fisher's exact test Fisher's exact test Fisher's exact test Fisher's exact test Fisher's exact test Fisher's exact test Fisher's exact test Fisher's exact test
BRAF 22 (10%) 202 9.13e-07
(0.000131) 		2.31e-08
(3.35e-06) 		4.17e-05
(0.00592) 		0.215
(1.00) 		0.304
(1.00) 		0.6
(1.00) 		1
(1.00)
KRAS 96 (43%) 128 7.45e-05
(0.0105) 		4.05e-05
(0.00579) 		0.485
(1.00) 		0.257
(1.00) 		0.634
(1.00) 		1
(1.00) 		0.0688
(1.00) 		0.735
(1.00)
TP53 120 (54%) 104 0.00202
(0.276) 		0.000482
(0.0668) 		7.28e-11
(1.06e-08) 		0.453
(1.00) 		0.782
(1.00) 		0.0252
(1.00) 		0.134
(1.00)
FBXW7 38 (17%) 186 0.0592
(1.00) 		0.0437
(1.00) 		0.000122
(0.0171) 		0.00881
(1.00) 		0.0362
(1.00) 		0.356
(1.00) 		0.118
(1.00)
PIK3CA 33 (15%) 191 0.205
(1.00) 		0.57
(1.00) 		0.00048
(0.0668) 		0.0876
(1.00) 		0.129
(1.00) 		0.0853
(1.00) 		1
(1.00)
APC 160 (71%) 64 0.0076
(1.00) 		0.0267
(1.00) 		0.221
(1.00) 		0.653
(1.00) 		0.452
(1.00) 		0.689
(1.00) 		0.464
(1.00)
NRAS 20 (9%) 204 0.0465
(1.00) 		0.125
(1.00) 		0.274
(1.00) 		0.247
(1.00) 		0.474
(1.00) 		1
(1.00) 		1
(1.00)
SMAD4 26 (12%) 198 0.0308
(1.00) 		0.288
(1.00) 		0.0355
(1.00) 		0.521
(1.00) 		0.339
(1.00) 		1
(1.00) 		0.846
(1.00)
FAM123B 25 (11%) 199 0.0723
(1.00) 		0.0918
(1.00) 		0.00539
(0.728) 		0.536
(1.00) 		0.588
(1.00) 		1
(1.00) 		0.459
(1.00)
SMAD2 15 (7%) 209 0.133
(1.00) 		0.518
(1.00) 		0.0393
(1.00) 		0.533
(1.00) 		0.133
(1.00) 		1
(1.00) 		0.711
(1.00)
TCF7L2 18 (8%) 206 1
(1.00) 		0.413
(1.00) 		0.417
(1.00) 		0.737
(1.00) 		0.896
(1.00) 		1
(1.00) 		0.23
(1.00)
ACVR2A 9 (4%) 215 0.00717
(0.961) 		0.184
(1.00) 		0.0325
(1.00) 		0.429
(1.00) 		0.469
(1.00) 		1
(1.00) 		0.545
(1.00)
SOX9 10 (4%) 214 0.129
(1.00) 		0.123
(1.00) 		0.00451
(0.613) 		0.204
(1.00) 		0.694
(1.00) 		1
(1.00) 		0.584
(1.00)
ELF3 6 (3%) 218 0.363
(1.00) 		0.811
(1.00) 		0.0514
(1.00) 		0.612
(1.00) 		0.436
(1.00) 		1
(1.00) 		0.406
(1.00)
CRTC1 6 (3%) 218 0.123
(1.00) 		0.165
(1.00) 		0.837
(1.00) 		0.209
(1.00) 		0.482
(1.00) 		1
(1.00) 		0.406
(1.00)
TNFRSF10C 6 (3%) 218 0.317
(1.00) 		0.287
(1.00) 		0.675
(1.00) 		0.853
(1.00) 		0.378
(1.00) 		1
(1.00) 		1
(1.00)
KRTAP5-5 4 (2%) 220 0.327
(1.00) 		0.482
(1.00) 		0.48
(1.00) 		0.388
(1.00) 		1
(1.00) 		1
(1.00) 		1
(1.00)
KIAA1804 15 (7%) 209 0.267
(1.00) 		0.397
(1.00) 		0.0659
(1.00) 		0.758
(1.00) 		0.474
(1.00) 		1
(1.00) 		0.736
(1.00)
ACOT4 3 (1%) 221 0.255
(1.00) 		0.49
(1.00) 		0.829
(1.00) 		1
(1.00) 		1
(1.00)
PTEN 7 (3%) 217 0.0276
(1.00) 		0.337
(1.00) 		0.0429
(1.00) 		0.612
(1.00) 		0.436
(1.00) 		0.244
(1.00) 		0.0575
(1.00)
MYO1B 13 (6%) 211 0.00862
(1.00) 		0.00879
(1.00) 		0.00752
(1.00) 		0.013
(1.00) 		0.469
(1.00) 		1
(1.00) 		0.653
(1.00)
'FBXW7 MUTATION STATUS' versus 'CN_CNMF'

P value = 0.000122 (Fisher's exact test), Q value = 0.017

Table S1.  Gene #2: 'FBXW7 MUTATION STATUS' versus Clinical Feature #3: 'CN_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 86 83 36 16
FBXW7 MUTATED 4 24 8 2
FBXW7 WILD-TYPE 82 59 28 14

Figure S1.  Get High-res Image Gene #2: 'FBXW7 MUTATION STATUS' versus Clinical Feature #3: 'CN_CNMF'

'BRAF MUTATION STATUS' versus 'MRNA_CNMF'

P value = 9.13e-07 (Fisher's exact test), Q value = 0.00013

Table S2.  Gene #4: 'BRAF MUTATION STATUS' versus Clinical Feature #1: 'MRNA_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 42 57 67 39
BRAF MUTATED 13 1 1 5
BRAF WILD-TYPE 29 56 66 34

Figure S2.  Get High-res Image Gene #4: 'BRAF MUTATION STATUS' versus Clinical Feature #1: 'MRNA_CNMF'

'BRAF MUTATION STATUS' versus 'MRNA_CHIERARCHICAL'

P value = 2.31e-08 (Fisher's exact test), Q value = 3.3e-06

Table S3.  Gene #4: 'BRAF MUTATION STATUS' versus Clinical Feature #2: 'MRNA_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 40 48 61 56
BRAF MUTATED 1 0 18 1
BRAF WILD-TYPE 39 48 43 55

Figure S3.  Get High-res Image Gene #4: 'BRAF MUTATION STATUS' versus Clinical Feature #2: 'MRNA_CHIERARCHICAL'

'BRAF MUTATION STATUS' versus 'CN_CNMF'

P value = 4.17e-05 (Fisher's exact test), Q value = 0.0059

Table S4.  Gene #4: 'BRAF MUTATION STATUS' versus Clinical Feature #3: 'CN_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 86 83 36 16
BRAF MUTATED 1 18 3 0
BRAF WILD-TYPE 85 65 33 16

Figure S4.  Get High-res Image Gene #4: 'BRAF MUTATION STATUS' versus Clinical Feature #3: 'CN_CNMF'

'KRAS MUTATION STATUS' versus 'MRNA_CNMF'

P value = 7.45e-05 (Fisher's exact test), Q value = 0.01

Table S5.  Gene #5: 'KRAS MUTATION STATUS' versus Clinical Feature #1: 'MRNA_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 42 57 67 39
KRAS MUTATED 13 14 31 27
KRAS WILD-TYPE 29 43 36 12

Figure S5.  Get High-res Image Gene #5: 'KRAS MUTATION STATUS' versus Clinical Feature #1: 'MRNA_CNMF'

'KRAS MUTATION STATUS' versus 'MRNA_CHIERARCHICAL'

P value = 4.05e-05 (Fisher's exact test), Q value = 0.0058

Table S6.  Gene #5: 'KRAS MUTATION STATUS' versus Clinical Feature #2: 'MRNA_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 40 48 61 56
KRAS MUTATED 10 33 26 16
KRAS WILD-TYPE 30 15 35 40

Figure S6.  Get High-res Image Gene #5: 'KRAS MUTATION STATUS' versus Clinical Feature #2: 'MRNA_CHIERARCHICAL'

'TP53 MUTATION STATUS' versus 'MRNA_CHIERARCHICAL'

P value = 0.000482 (Fisher's exact test), Q value = 0.067

Table S7.  Gene #6: 'TP53 MUTATION STATUS' versus Clinical Feature #2: 'MRNA_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 40 48 61 56
TP53 MUTATED 30 19 25 36
TP53 WILD-TYPE 10 29 36 20

Figure S7.  Get High-res Image Gene #6: 'TP53 MUTATION STATUS' versus Clinical Feature #2: 'MRNA_CHIERARCHICAL'

'TP53 MUTATION STATUS' versus 'CN_CNMF'

P value = 7.28e-11 (Fisher's exact test), Q value = 1.1e-08

Table S8.  Gene #6: 'TP53 MUTATION STATUS' versus Clinical Feature #3: 'CN_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 86 83 36 16
TP53 MUTATED 68 22 19 9
TP53 WILD-TYPE 18 61 17 7

Figure S8.  Get High-res Image Gene #6: 'TP53 MUTATION STATUS' versus Clinical Feature #3: 'CN_CNMF'

'PIK3CA MUTATION STATUS' versus 'CN_CNMF'

P value = 0.00048 (Fisher's exact test), Q value = 0.067

Table S9.  Gene #9: 'PIK3CA MUTATION STATUS' versus Clinical Feature #3: 'CN_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 86 83 36 16
PIK3CA MUTATED 5 22 2 3
PIK3CA WILD-TYPE 81 61 34 13

Figure S9.  Get High-res Image Gene #9: 'PIK3CA MUTATION STATUS' versus Clinical Feature #3: 'CN_CNMF'

Methods & Data
Input

  • Mutation data file = COADREAD-TP.mutsig.cluster.txt

  • Molecular subtypes file = COADREAD-TP.transferedmergedcluster.txt

  • Number of patients = 224

  • Number of significantly mutated genes = 21

  • Number of Molecular subtypes = 8

  • Exclude genes that fewer than K tumors have mutations, K = 3

Fisher's exact test

For binary or multi-class clinical features (nominal or ordinal), 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.

References
[1] 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)
[2] 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