A novel independence test for somatic alterations in cancer shows that biology drives mutual exclusivity but chance explains most co-occurrence

Abstract

In cancer, mutually exclusive or co-occurring somatic alterations across genes can suggest functional interactions. Existing tests for such patterns make the unrealistic assumption of identical gene alteration probabilities across tumors. We present Discrete Independence Statistic Controlling for Observations with Varying Event Rates (DISCOVER), a novel test that is more sensitive than other methods and controls its false positive rate. A pan-cancer analysis using DISCOVER finds no evidence for widespread co-occurrence, and most co-occurrences previously detected do not exceed expectation by chance. Many mutual exclusivities are identified involving well-known genes related to cell cycle and growth factor signaling, as well as lesser known regulators of Hedgehog signaling.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-016-1114-x) contains supplementary material, which is available to authorized users.

Mapping the in vivo fitness landscape of lung adenocarcinoma tumor suppression in mice

“…Deletions from TCGA datasets were called as described in Curtis et al 31 . The DISCOVER statistical test was used to determine co-occurrence and mutual exclusivity 13 . A total of 230 samples (TCGA, 2014) 1 and 1,563 samples (GENIE, 2017) 4 were studied.…”

“…4e,f) 1,4 . Notably, despite a ~5-fold enrichment in the co-occurrence of these two alterations, this interaction would be statistically insignificant in a naive survey of all potential pairwise driver interactions after correcting for multiple-hypothesis testing, thus illustrating the need to study genetic interactions beyond co-occurrence patterns ( P = 0.32 after Bonferroni correction for 10 pairwise interactions) 13 .…”

“…Even the tumor suppressor alterations that conferred advantage in all three contexts ( Rb1 and Apc ) still exhibited context-dependent magnitudes of tumor suppression. Such widespread context dependency is overlooked by global surveys of drivers, where driver interactions are either ignored 1,2 or presumed to be sufficiently rare and/or weak to justify considering only marginal correlations 3,13 . Nonetheless, our fitness measurements generally agree with mutation co-occurrence patterns in human lung cancer, despite the limited statistical resolution of these data (Spearman R = 0.50, P = 0.03, Supplementary Fig.…”

A p53 Super-tumor Suppressor Reveals a Tumor Suppressive p53-Ptpn14-Yap Axis in Pancreatic Cancer

“…The DISCOVER package, which employs proper false positive rate controls in comparison to the Fisher test (Canisius et al, 2016), was used to test for mutual exclusivity. Mutual exclusivity plots were generated with Bioconductor package ComplexHeatmap (Gu et al, 2016).…”

Survival Outcomes by TP53 Mutation Status in Metastatic Breast Cancer

“…DISCOVER, a statistical test for detecting co-occurrence and mutual exclusivity in cancer genomics data was used. 8 Unlike traditional approaches such as Fisher’s exact test, DISCOVER is based on a null model that takes into account the overall tumor-specific alteration rates when deciding whether alterations co-occur more or less often than expected by chance. Multiple testing was adjusted using false discovery rate (FDR).…”