Correlation between gene mutation status and molecular subtypes
Thyroid Adenocarcinoma (HistologicalType_Follicular)
01 July 2013  |  awg_thca__2013_07_01
Maintainer Information
Citation Information
Maintained by TCGA GDAC Team (Broad Institute/MD Anderson Cancer Center/Harvard Medical School)
Cite as Broad Institute TCGA Genome Data Analysis Center (2013): Correlation between gene mutation status and molecular subtypes. Broad Institute of MIT and Harvard. doi:10.7908/C14F1NWP
Overview
Introduction

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

Summary

Testing the association between mutation status of 3 genes and 9 molecular subtypes across 83 patients, 15 significant findings detected with P value < 0.05 and Q value < 0.25.

  • NRAS mutation correlated to 'METHLYATION_CNMF',  'MRNASEQ_CNMF',  'MRNASEQ_CHIERARCHICAL', and 'MIRSEQ_CHIERARCHICAL'.

  • BRAF mutation correlated to 'METHLYATION_CNMF',  'MRNASEQ_CNMF',  'MRNASEQ_CHIERARCHICAL',  'MIRSEQ_CNMF',  'MIRSEQ_CHIERARCHICAL',  'MIRSEQ_MATURE_CNMF', and 'MIRSEQ_MATURE_CHIERARCHICAL'.

  • HRAS mutation correlated to 'METHLYATION_CNMF',  'MRNASEQ_CNMF',  'MRNASEQ_CHIERARCHICAL', and 'MIRSEQ_MATURE_CHIERARCHICAL'.

Results
Overview of the results

Table 1.  Get Full Table Overview of the association between mutation status of 3 genes and 9 molecular subtypes. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, 15 significant findings detected.

Clinical
Features
METHLYATION
CNMF
RPPA
CNMF
RPPA
CHIERARCHICAL
MRNASEQ
CNMF
MRNASEQ
CHIERARCHICAL
MIRSEQ
CNMF
MIRSEQ
CHIERARCHICAL
MIRSEQ
MATURE
CNMF
MIRSEQ
MATURE
CHIERARCHICAL
nMutated (%) nWild-Type Chi-square 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 Fisher's exact test
BRAF 13 (16%) 70 1.67e-06
(4.17e-05)
1
(1.00)
0.698
(1.00)
1.35e-07
(3.52e-06)
4.3e-09
(1.16e-07)
1.46e-05
(0.00032)
1.73e-06
(4.17e-05)
0.000763
(0.013)
7.85e-05
(0.00157)
NRAS 19 (23%) 64 9.1e-06
(0.000209)
0.585
(1.00)
0.247
(1.00)
0.00155
(0.0248)
2.5e-05
(0.000525)
0.498
(1.00)
0.00909
(0.132)
0.142
(1.00)
1
(1.00)
HRAS 8 (10%) 75 0.000225
(0.00404)
0.0645
(0.71)
0.399
(1.00)
0.0109
(0.142)
0.000139
(0.00264)
0.0296
(0.356)
1
(1.00)
0.131
(1.00)
0.00878
(0.132)
'NRAS MUTATION STATUS' versus 'METHLYATION_CNMF'

P value = 9.1e-06 (Chi-square test), Q value = 0.00021

Table S1.  Gene #1: 'NRAS MUTATION STATUS' versus Clinical Feature #1: 'METHLYATION_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4 CLUS_5 CLUS_6 CLUS_7 CLUS_8
ALL 7 18 16 1 10 8 17 6
NRAS MUTATED 3 0 8 1 1 0 1 5
NRAS WILD-TYPE 4 18 8 0 9 8 16 1

Figure S1.  Get High-res Image Gene #1: 'NRAS MUTATION STATUS' versus Clinical Feature #1: 'METHLYATION_CNMF'

'NRAS MUTATION STATUS' versus 'MRNASEQ_CNMF'

P value = 0.00155 (Fisher's exact test), Q value = 0.025

Table S2.  Gene #1: 'NRAS MUTATION STATUS' versus Clinical Feature #4: 'MRNASEQ_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 21 17 22 23
NRAS MUTATED 11 2 1 5
NRAS WILD-TYPE 10 15 21 18

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

'NRAS MUTATION STATUS' versus 'MRNASEQ_CHIERARCHICAL'

P value = 2.5e-05 (Fisher's exact test), Q value = 0.00053

Table S3.  Gene #1: 'NRAS MUTATION STATUS' versus Clinical Feature #5: 'MRNASEQ_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 15 18 25 25
NRAS MUTATED 0 0 6 13
NRAS WILD-TYPE 15 18 19 12

Figure S3.  Get High-res Image Gene #1: 'NRAS MUTATION STATUS' versus Clinical Feature #5: 'MRNASEQ_CHIERARCHICAL'

'NRAS MUTATION STATUS' versus 'MIRSEQ_CHIERARCHICAL'

P value = 0.00909 (Fisher's exact test), Q value = 0.13

Table S4.  Gene #1: 'NRAS MUTATION STATUS' versus Clinical Feature #7: 'MIRSEQ_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2
ALL 17 66
NRAS MUTATED 0 19
NRAS WILD-TYPE 17 47

Figure S4.  Get High-res Image Gene #1: 'NRAS MUTATION STATUS' versus Clinical Feature #7: 'MIRSEQ_CHIERARCHICAL'

'BRAF MUTATION STATUS' versus 'METHLYATION_CNMF'

P value = 1.67e-06 (Chi-square test), Q value = 4.2e-05

Table S5.  Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #1: 'METHLYATION_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4 CLUS_5 CLUS_6 CLUS_7 CLUS_8
ALL 7 18 16 1 10 8 17 6
BRAF MUTATED 0 1 1 0 0 0 11 0
BRAF WILD-TYPE 7 17 15 1 10 8 6 6

Figure S5.  Get High-res Image Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #1: 'METHLYATION_CNMF'

'BRAF MUTATION STATUS' versus 'MRNASEQ_CNMF'

P value = 1.35e-07 (Fisher's exact test), Q value = 3.5e-06

Table S6.  Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #4: 'MRNASEQ_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 21 17 22 23
BRAF MUTATED 1 0 12 0
BRAF WILD-TYPE 20 17 10 23

Figure S6.  Get High-res Image Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #4: 'MRNASEQ_CNMF'

'BRAF MUTATION STATUS' versus 'MRNASEQ_CHIERARCHICAL'

P value = 4.3e-09 (Fisher's exact test), Q value = 1.2e-07

Table S7.  Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #5: 'MRNASEQ_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 15 18 25 25
BRAF MUTATED 0 12 0 1
BRAF WILD-TYPE 15 6 25 24

Figure S7.  Get High-res Image Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #5: 'MRNASEQ_CHIERARCHICAL'

'BRAF MUTATION STATUS' versus 'MIRSEQ_CNMF'

P value = 1.46e-05 (Fisher's exact test), Q value = 0.00032

Table S8.  Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #6: 'MIRSEQ_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3
ALL 38 25 20
BRAF MUTATED 0 4 9
BRAF WILD-TYPE 38 21 11

Figure S8.  Get High-res Image Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #6: 'MIRSEQ_CNMF'

'BRAF MUTATION STATUS' versus 'MIRSEQ_CHIERARCHICAL'

P value = 1.73e-06 (Fisher's exact test), Q value = 4.2e-05

Table S9.  Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #7: 'MIRSEQ_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2
ALL 17 66
BRAF MUTATED 10 3
BRAF WILD-TYPE 7 63

Figure S9.  Get High-res Image Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #7: 'MIRSEQ_CHIERARCHICAL'

'BRAF MUTATION STATUS' versus 'MIRSEQ_MATURE_CNMF'

P value = 0.000763 (Fisher's exact test), Q value = 0.013

Table S10.  Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #8: 'MIRSEQ_MATURE_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 23 18 25 17
BRAF MUTATED 1 3 1 8
BRAF WILD-TYPE 22 15 24 9

Figure S10.  Get High-res Image Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #8: 'MIRSEQ_MATURE_CNMF'

'BRAF MUTATION STATUS' versus 'MIRSEQ_MATURE_CHIERARCHICAL'

P value = 7.85e-05 (Fisher's exact test), Q value = 0.0016

Table S11.  Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #9: 'MIRSEQ_MATURE_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2
ALL 48 35
BRAF MUTATED 1 12
BRAF WILD-TYPE 47 23

Figure S11.  Get High-res Image Gene #2: 'BRAF MUTATION STATUS' versus Clinical Feature #9: 'MIRSEQ_MATURE_CHIERARCHICAL'

'HRAS MUTATION STATUS' versus 'METHLYATION_CNMF'

P value = 0.000225 (Chi-square test), Q value = 0.004

Table S12.  Gene #3: 'HRAS MUTATION STATUS' versus Clinical Feature #1: 'METHLYATION_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4 CLUS_5 CLUS_6 CLUS_7 CLUS_8
ALL 7 18 16 1 10 8 17 6
HRAS MUTATED 1 0 7 0 0 0 0 0
HRAS WILD-TYPE 6 18 9 1 10 8 17 6

Figure S12.  Get High-res Image Gene #3: 'HRAS MUTATION STATUS' versus Clinical Feature #1: 'METHLYATION_CNMF'

'HRAS MUTATION STATUS' versus 'MRNASEQ_CNMF'

P value = 0.0109 (Fisher's exact test), Q value = 0.14

Table S13.  Gene #3: 'HRAS MUTATION STATUS' versus Clinical Feature #4: 'MRNASEQ_CNMF'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 21 17 22 23
HRAS MUTATED 6 0 1 1
HRAS WILD-TYPE 15 17 21 22

Figure S13.  Get High-res Image Gene #3: 'HRAS MUTATION STATUS' versus Clinical Feature #4: 'MRNASEQ_CNMF'

'HRAS MUTATION STATUS' versus 'MRNASEQ_CHIERARCHICAL'

P value = 0.000139 (Fisher's exact test), Q value = 0.0026

Table S14.  Gene #3: 'HRAS MUTATION STATUS' versus Clinical Feature #5: 'MRNASEQ_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2 CLUS_3 CLUS_4
ALL 15 18 25 25
HRAS MUTATED 0 0 0 8
HRAS WILD-TYPE 15 18 25 17

Figure S14.  Get High-res Image Gene #3: 'HRAS MUTATION STATUS' versus Clinical Feature #5: 'MRNASEQ_CHIERARCHICAL'

'HRAS MUTATION STATUS' versus 'MIRSEQ_MATURE_CHIERARCHICAL'

P value = 0.00878 (Fisher's exact test), Q value = 0.13

Table S15.  Gene #3: 'HRAS MUTATION STATUS' versus Clinical Feature #9: 'MIRSEQ_MATURE_CHIERARCHICAL'

nPatients CLUS_1 CLUS_2
ALL 48 35
HRAS MUTATED 1 7
HRAS WILD-TYPE 47 28

Figure S15.  Get High-res Image Gene #3: 'HRAS MUTATION STATUS' versus Clinical Feature #9: 'MIRSEQ_MATURE_CHIERARCHICAL'

Methods & Data
Input
  • Mutation data file = THCA-HistologicalType_Follicular.mutsig.cluster.txt

  • Molecular subtypes file = THCA-HistologicalType_Follicular.transferedmergedcluster.txt

  • Number of patients = 83

  • Number of significantly mutated genes = 3

  • Number of Molecular subtypes = 9

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

Chi-square test

For multi-class clinical features (nominal or ordinal), Chi-square tests (Greenwood and Nikulin 1996) were used to estimate the P values using the 'chisq.test' function in R

Fisher's exact test

For binary 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.

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] Greenwood and Nikulin, A guide to chi-squared testing, Wiley, New York. ISBN 047155779X (1996)
[2] 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)
[3] 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)