Correlation between gene mutation status and selected clinical features
Acute Myeloid Leukemia (Primary blood derived cancer - Peripheral blood)
28 January 2016  |  analyses__2016_01_28
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 (2016): Correlation between gene mutation status and selected clinical features. Broad Institute of MIT and Harvard. doi:10.7908/C1JQ10FG
Overview
Introduction

This pipeline computes the correlation between significantly recurrent gene mutations and selected clinical features.

Summary

Testing the association between mutation status of 23 genes and 5 clinical features across 193 patients, 7 significant findings detected with Q value < 0.25.

  • DNMT3A mutation correlated to 'Time to Death'.

  • IDH2 mutation correlated to 'YEARS_TO_BIRTH'.

  • U2AF1 mutation correlated to 'YEARS_TO_BIRTH'.

  • RUNX1 mutation correlated to 'YEARS_TO_BIRTH'.

  • CEBPA mutation correlated to 'YEARS_TO_BIRTH'.

  • TP53 mutation correlated to 'Time to Death' and 'YEARS_TO_BIRTH'.

Results
Overview of the results

Table 1.  Get Full Table Overview of the association between mutation status of 23 genes and 5 clinical features. Shown in the table are P values (Q values). Thresholded by Q value < 0.25, 7 significant findings detected.

Clinical
Features
Time
to
Death
YEARS
TO
BIRTH
GENDER RACE ETHNICITY
nMutated (%) nWild-Type logrank test Wilcoxon-test Fisher's exact test Fisher's exact test Fisher's exact test
TP53 15 (8%) 178 5.68e-06
(0.000653)
0.000713
(0.041)
0.176
(0.724)
1
(1.00)
1
(1.00)
DNMT3A 48 (25%) 145 0.0011
(0.042)
0.267
(0.837)
0.136
(0.724)
0.382
(0.935)
0.158
(0.724)
IDH2 20 (10%) 173 0.685
(1.00)
0.00164
(0.0471)
0.815
(1.00)
0.36
(0.9)
0.285
(0.86)
U2AF1 8 (4%) 185 0.47
(1.00)
0.00422
(0.097)
0.0696
(0.616)
1
(1.00)
1
(1.00)
RUNX1 16 (8%) 177 0.0284
(0.356)
0.00656
(0.126)
1
(1.00)
1
(1.00)
1
(1.00)
CEBPA 13 (7%) 180 0.765
(1.00)
0.0117
(0.193)
0.58
(1.00)
1
(1.00)
1
(1.00)
FLT3 52 (27%) 141 0.167
(0.724)
0.331
(0.865)
0.627
(1.00)
0.0693
(0.616)
0.563
(1.00)
NPM1 33 (17%) 160 0.705
(1.00)
0.172
(0.724)
0.57
(1.00)
0.522
(1.00)
1
(1.00)
IDH1 18 (9%) 175 0.638
(1.00)
0.326
(0.865)
0.465
(1.00)
0.701
(1.00)
1
(1.00)
TET2 17 (9%) 176 0.844
(1.00)
0.127
(0.724)
0.319
(0.865)
1
(1.00)
1
(1.00)
NRAS 15 (8%) 178 0.666
(1.00)
0.256
(0.837)
1
(1.00)
0.67
(1.00)
1
(1.00)
WT1 12 (6%) 181 0.574
(1.00)
0.168
(0.724)
0.774
(1.00)
0.648
(1.00)
0.165
(0.724)
KRAS 8 (4%) 185 0.303
(0.861)
0.102
(0.724)
0.149
(0.724)
1
(1.00)
0.122
(0.724)
PHF6 6 (3%) 187 0.977
(1.00)
0.158
(0.724)
0.0309
(0.356)
1
(1.00)
1
(1.00)
STAG2 6 (3%) 187 0.34
(0.87)
0.215
(0.808)
0.42
(1.00)
0.13
(0.724)
1
(1.00)
KIT 8 (4%) 185 0.489
(1.00)
0.681
(1.00)
0.476
(1.00)
1
(1.00)
1
(1.00)
RAD21 5 (3%) 188 0.888
(1.00)
0.233
(0.837)
1
(1.00)
0.0191
(0.275)
1
(1.00)
EZH2 3 (2%) 190 0.213
(0.808)
1
(1.00)
1
(1.00)
1
(1.00)
SMC3 7 (4%) 186 0.117
(0.724)
0.783
(1.00)
0.707
(1.00)
1
(1.00)
1
(1.00)
ASXL1 5 (3%) 188 0.269
(0.837)
0.0661
(0.616)
0.666
(1.00)
1
(1.00)
0.0773
(0.635)
SMC1A 6 (3%) 187 0.307
(0.861)
0.265
(0.837)
0.218
(0.808)
0.435
(1.00)
1
(1.00)
PTPN11 9 (5%) 184 0.554
(1.00)
0.292
(0.86)
1
(1.00)
0.576
(1.00)
1
(1.00)
SUZ12 3 (2%) 190 0.88
(1.00)
0.946
(1.00)
0.25
(0.837)
1
(1.00)
'DNMT3A MUTATION STATUS' versus 'Time to Death'

P value = 0.0011 (logrank test), Q value = 0.042

Table S1.  Gene #2: 'DNMT3A MUTATION STATUS' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 179 115 0.0 - 94.1 (12.0)
DNMT3A MUTATED 44 34 0.0 - 34.0 (8.5)
DNMT3A WILD-TYPE 135 81 0.0 - 94.1 (13.0)

Figure S1.  Get High-res Image Gene #2: 'DNMT3A MUTATION STATUS' versus Clinical Feature #1: 'Time to Death'

'IDH2 MUTATION STATUS' versus 'YEARS_TO_BIRTH'

P value = 0.00164 (Wilcoxon-test), Q value = 0.047

Table S2.  Gene #4: 'IDH2 MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

nPatients Mean (Std.Dev)
ALL 193 54.9 (16.2)
IDH2 MUTATED 20 64.9 (8.0)
IDH2 WILD-TYPE 173 53.8 (16.5)

Figure S2.  Get High-res Image Gene #4: 'IDH2 MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'U2AF1 MUTATION STATUS' versus 'YEARS_TO_BIRTH'

P value = 0.00422 (Wilcoxon-test), Q value = 0.097

Table S3.  Gene #9: 'U2AF1 MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

nPatients Mean (Std.Dev)
ALL 193 54.9 (16.2)
U2AF1 MUTATED 8 69.9 (9.0)
U2AF1 WILD-TYPE 185 54.3 (16.1)

Figure S3.  Get High-res Image Gene #9: 'U2AF1 MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RUNX1 MUTATION STATUS' versus 'YEARS_TO_BIRTH'

P value = 0.00656 (Wilcoxon-test), Q value = 0.13

Table S4.  Gene #10: 'RUNX1 MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

nPatients Mean (Std.Dev)
ALL 193 54.9 (16.2)
RUNX1 MUTATED 16 64.3 (16.0)
RUNX1 WILD-TYPE 177 54.1 (16.0)

Figure S4.  Get High-res Image Gene #10: 'RUNX1 MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'CEBPA MUTATION STATUS' versus 'YEARS_TO_BIRTH'

P value = 0.0117 (Wilcoxon-test), Q value = 0.19

Table S5.  Gene #11: 'CEBPA MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

nPatients Mean (Std.Dev)
ALL 193 54.9 (16.2)
CEBPA MUTATED 13 42.7 (17.6)
CEBPA WILD-TYPE 180 55.8 (15.8)

Figure S5.  Get High-res Image Gene #11: 'CEBPA MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'TP53 MUTATION STATUS' versus 'Time to Death'

P value = 5.68e-06 (logrank test), Q value = 0.00065

Table S6.  Gene #12: 'TP53 MUTATION STATUS' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 179 115 0.0 - 94.1 (12.0)
TP53 MUTATED 14 14 0.0 - 17.0 (6.0)
TP53 WILD-TYPE 165 101 0.0 - 94.1 (12.0)

Figure S6.  Get High-res Image Gene #12: 'TP53 MUTATION STATUS' versus Clinical Feature #1: 'Time to Death'

'TP53 MUTATION STATUS' versus 'YEARS_TO_BIRTH'

P value = 0.000713 (Wilcoxon-test), Q value = 0.041

Table S7.  Gene #12: 'TP53 MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

nPatients Mean (Std.Dev)
ALL 193 54.9 (16.2)
TP53 MUTATED 15 67.8 (10.1)
TP53 WILD-TYPE 178 53.8 (16.2)

Figure S7.  Get High-res Image Gene #12: 'TP53 MUTATION STATUS' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

Methods & Data
Input
  • Mutation data file = sample_sig_gene_table.txt from Mutsig_2CV pipeline

  • Processed Mutation data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/GDAC_Correlate_Genomic_Events_Preprocess/LAML-TB/22570966/transformed.cor.cli.txt

  • Clinical data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/Append_Data/LAML-TB/22506504/LAML-TB.merged_data.txt

  • Number of patients = 193

  • Number of significantly mutated genes = 23

  • Number of selected clinical features = 5

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

Survival analysis

For survival clinical features, the Kaplan-Meier survival curves of tumors with and without gene mutations were plotted and the statistical significance P values were estimated by logrank test (Bland and Altman 2004) using the 'survdiff' function in R

Fisher's exact test

For binary 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

In addition to the links below, the full results of the analysis summarized in this report can also be downloaded programmatically using firehose_get, or interactively from either the Broad GDAC website or TCGA Data Coordination Center Portal.

References
[1] Bland and Altman, Statistics notes: The logrank test, BMJ 328(7447):1073 (2004)
[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)