Identification of cis-regulatory mutations generating de novo edges in personalized cancer gene regulatory networks

Zeynep Kalender Atak,Hugo Ceulemans,Dmitry Svetlichnyy,Gert Hulselmans,Stein Aerts,Joke Reumers,Hana Imrichova,Valerie Christiaens

doi:10.1186/s13073-017-0464-7

Zeynep Kalender Atak, Hugo Ceulemans + Show 6 more

Open Access

https://doi.org/10.1186/s13073-017-0464-7

Copy DOI

Abstract

The identification of functional non-coding mutations is a key challenge in the field of genomics. Here we introduce μ-cisTarget to filter, annotate and prioritize cis-regulatory mutations based on their putative effect on the underlying “personal” gene regulatory network. We validated μ-cisTarget by re-analyzing the TAL1 and LMO1 enhancer mutations in T-ALL, and the TERT promoter mutation in melanoma. Next, we re-sequenced the full genomes of ten cancer cell lines and used matched transcriptome data and motif discovery to identify master regulators with de novo binding sites that result in the up-regulation of nearby oncogenic drivers. μ-cisTarget is available from http://mucistarget.aertslab.org.

Highlights

Oncogenic programs are characterized by aberrant gene expression profiles
A small number of non-coding mutations generate de novo oncogenic edges in driver gene regulatory networks We developed a new computational pipeline, called μcisTarget, to identify cis-regulatory mutations in a cancer sample, when both the whole genome sequence and the gene expression profile of that sample are available
Among the list of candidates, we further determine a final set of mutations by applying two filters, namely: (i) the transcription factor for which a binding site is generated is itself expressed in the sample and is related to the cancer type; and (ii) the mutation is located close to a target gene that is overexpressed, and within the same topologically associating domain (TAD) as the over-expressed target, it is related to the cancer type under study, and/or it is a potential driver gene (Fig. 1; see “Methods”)

Summary

Introduction

Oncogenic programs are characterized by aberrant gene expression profiles. A gene regulatory network underlying a cancer transcriptome can be considered as a perturbed stable network configuration, or as a cancer attractor state [1]. Gene expression changes leading from a normal cell to a malignant state are generally due to a series of acquired somatic mutations, which often affect proteins playing a key role in transcriptional regulation [2]. These can include mutations, amplifications, or translocations leading to an altered function or expression of transcription factors (e.g., MYC, TAL1, MITF, TP53), co-factors (EZH2, RB1, IDH1, MLL), or signaling molecules that lead to downstream alterations in transcription factor activity (e.g., RAS/RAC/RAF, KIT, PTEN, CDKN2A). More subtle changes can occur in gene regulatory networks, which may cause finetuning of the emerging transcriptome rather than necessarily yielding a different attractor state.

Methods

Results

Discussion

Conclusion