Correlation between copy number variations of arm-level result and selected clinical features
Acute Myeloid Leukemia (Primary blood derived cancer - Peripheral blood)
15 July 2014  |  analyses__2014_07_15
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 (2014): Correlation between copy number variations of arm-level result and selected clinical features. Broad Institute of MIT and Harvard. doi:10.7908/C1N878JV
Overview
Introduction

This pipeline computes the correlation between significant arm-level copy number variations (cnvs) and selected clinical features.

Summary

Testing the association between copy number variation 30 arm-level events and 5 clinical features across 191 patients, 2 significant findings detected with Q value < 0.25.

  • 5q loss cnv correlated to 'Time to Death'.

  • 18q loss cnv correlated to 'Time to Death'.

Results
Overview of the results

Table 1.  Get Full Table Overview of the association between significant copy number variation of 30 arm-level events and 5 clinical features. Shown in the table are P values (Q values). Thresholded by Q value < 0.25, 2 significant findings detected.

Clinical
Features
Time
to
Death
AGE GENDER RACE ETHNICITY
nCNV (%) nWild-Type logrank test Wilcoxon-test Fisher's exact test Fisher's exact test Fisher's exact test
5q loss 6 (3%) 185 0.000704
(0.101)
0.0519
(1.00)
0.0325
(1.00)
0.416
(1.00)
1
(1.00)
18q loss 4 (2%) 187 0.000836
(0.119)
0.0121
(1.00)
0.127
(1.00)
1
(1.00)
1
(1.00)
1p gain 3 (2%) 188 0.115
(1.00)
0.592
(1.00)
1
(1.00)
1
(1.00)
4p gain 4 (2%) 187 0.583
(1.00)
0.328
(1.00)
0.627
(1.00)
1
(1.00)
1
(1.00)
4q gain 4 (2%) 187 0.583
(1.00)
0.328
(1.00)
0.627
(1.00)
1
(1.00)
1
(1.00)
8p gain 22 (12%) 169 0.514
(1.00)
0.112
(1.00)
0.0734
(1.00)
1
(1.00)
0.313
(1.00)
8q gain 23 (12%) 168 0.56
(1.00)
0.167
(1.00)
0.0723
(1.00)
1
(1.00)
0.325
(1.00)
10q gain 3 (2%) 188 0.654
(1.00)
0.252
(1.00)
0.233
(1.00)
1
(1.00)
11p gain 4 (2%) 187 0.173
(1.00)
1
(1.00)
1
(1.00)
0.0628
(1.00)
11q gain 7 (4%) 184 0.566
(1.00)
0.047
(1.00)
1
(1.00)
1
(1.00)
0.108
(1.00)
13q gain 6 (3%) 185 0.826
(1.00)
0.12
(1.00)
0.69
(1.00)
1
(1.00)
1
(1.00)
17q gain 3 (2%) 188 0.737
(1.00)
0.113
(1.00)
1
(1.00)
1
(1.00)
1
(1.00)
19p gain 5 (3%) 186 0.741
(1.00)
0.715
(1.00)
0.378
(1.00)
1
(1.00)
1
(1.00)
19q gain 5 (3%) 186 0.741
(1.00)
0.715
(1.00)
0.378
(1.00)
1
(1.00)
1
(1.00)
21q gain 8 (4%) 183 0.0186
(1.00)
0.401
(1.00)
0.0732
(1.00)
0.186
(1.00)
1
(1.00)
22q gain 9 (5%) 182 0.728
(1.00)
0.0629
(1.00)
0.513
(1.00)
1
(1.00)
1
(1.00)
xq gain 3 (2%) 188 0.108
(1.00)
1
(1.00)
1
(1.00)
1
(1.00)
3p loss 3 (2%) 188 0.0208
(1.00)
0.252
(1.00)
1
(1.00)
1
(1.00)
3q loss 3 (2%) 188 0.0208
(1.00)
0.252
(1.00)
1
(1.00)
1
(1.00)
7p loss 17 (9%) 174 0.0328
(1.00)
0.458
(1.00)
0.802
(1.00)
1
(1.00)
1
(1.00)
7q loss 20 (10%) 171 0.015
(1.00)
0.513
(1.00)
1
(1.00)
0.718
(1.00)
1
(1.00)
12p loss 4 (2%) 187 0.869
(1.00)
0.627
(1.00)
1
(1.00)
1
(1.00)
15q loss 4 (2%) 187 0.679
(1.00)
0.677
(1.00)
1
(1.00)
1
(1.00)
1
(1.00)
16q loss 4 (2%) 187 0.0589
(1.00)
0.388
(1.00)
0.627
(1.00)
1
(1.00)
1
(1.00)
17p loss 13 (7%) 178 0.109
(1.00)
0.765
(1.00)
0.147
(1.00)
0.655
(1.00)
1
(1.00)
17q loss 7 (4%) 184 0.403
(1.00)
0.196
(1.00)
0.458
(1.00)
1
(1.00)
1
(1.00)
18p loss 5 (3%) 186 0.00203
(0.287)
0.367
(1.00)
0.0642
(1.00)
0.362
(1.00)
1
(1.00)
19p loss 4 (2%) 187 0.0991
(1.00)
0.0472
(1.00)
1
(1.00)
1
(1.00)
1
(1.00)
19q loss 4 (2%) 187 0.0991
(1.00)
0.0472
(1.00)
1
(1.00)
1
(1.00)
1
(1.00)
xq loss 5 (3%) 186 0.181
(1.00)
0.0283
(1.00)
0.0642
(1.00)
1
(1.00)
1
(1.00)
'5q loss' versus 'Time to Death'

P value = 0.000704 (logrank test), Q value = 0.1

Table S1.  Gene #18: '5q loss' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 168 106 0.9 - 94.1 (12.0)
5Q LOSS MUTATED 6 6 1.0 - 12.0 (7.0)
5Q LOSS WILD-TYPE 162 100 0.9 - 94.1 (12.5)

Figure S1.  Get High-res Image Gene #18: '5q loss' versus Clinical Feature #1: 'Time to Death'

'18q loss' versus 'Time to Death'

P value = 0.000836 (logrank test), Q value = 0.12

Table S2.  Gene #27: '18q loss' versus Clinical Feature #1: 'Time to Death'

nPatients nDeath Duration Range (Median), Month
ALL 168 106 0.9 - 94.1 (12.0)
18Q LOSS MUTATED 4 4 1.0 - 10.0 (4.5)
18Q LOSS WILD-TYPE 164 102 0.9 - 94.1 (12.0)

Figure S2.  Get High-res Image Gene #27: '18q loss' versus Clinical Feature #1: 'Time to Death'

Methods & Data
Input
  • Copy number data file = transformed.cor.cli.txt

  • Clinical data file = LAML-TB.merged_data.txt

  • Number of patients = 191

  • Number of significantly arm-level cnvs = 30

  • Number of selected clinical features = 5

  • Exclude regions 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)