DriveWays: A Method for Identifying Possibly Overlapping Driver Pathways in Cancer

Abstract

Motivation: The majority of the previous methods for identifying cancer driver modules output non-overlapping modules. This assumption is biologically inaccurate as genes can participate in multiple molecular pathways. This is particularly true for cancer-associated genes as many of them are network hubs connecting functionally distinct set of genes. It is important to provide combinatorial optimization problem definitions modeling this biological phenomenon and to suggest efficient algorithms for its solution. Results: We provide a formal definition of the Overlapping Driver Module Identification in Cancer (ODMIC) problem. We show that the problem is NP-hard. We propose a seed-and-extend based heuristic named DriveWays that identifies overlapping cancer driver modules from the graph built from the IntAct PPI network. DriveWays incorporates mutual exclusivity, coverage, and the network connectivity information of the genes. We show that DriveWays outperforms the state-of-the-art methods in recovering well-known cancer driver genes performed on TCGA pan-cancer data. Additionally, DriveWays's output modules show a stronger enrichment for the reference pathways in almost all cases. Overall, we show that enabling modules to overlap improves the recovery of functional pathways filtered with known cancer drivers, which essentially constitute the reference set of cancer-related pathways. Availability: The data, the source code, and useful scripts are available at: https://github.com/ abu-compbio/DriveWays Supplementary information: Supplementary data are available at Biorxiv.Recent advances in high-throughput DNA sequencing technology have allowed several projects such as The Cancer Genome Atlas (TCGA) to systematically generate genomic data for thousands of tumors across many cancer types [1].A key fundamental challenge in cancer genomics is to distinguish functional mutations that drive tumorigenesis, or drivers, from the numerous non-functional passenger mutations that occur randomly but that are not important for cancer development. Such a challenge is further complicated by the highly interactive nature of genes/proteins, thus necessitating the identification not only of such drivers but also of modules consisting of webs of drivers culpable in

A Network-centric Framework for the Evaluation of Mutual Exclusivity Tests on Cancer Drivers

“…A phenomenon observed frequently in the data pertaining to the alterations that the tumors acquire is mutual exclusivity (ME) ; a driver mutation is less likely to occur in case of an earlier mutation that has common functionality in the same molecular pathway [Leiserson et al, 2016, van de Haar et al, 2019, Thomas et al, 2007, Yeang and Levine, 2008]. Therefore several driver gene or module identification approaches employ ME detection as part of their problem definitions and optimization goals [Babur et al, 2015, Ciriello et al, 2012, Leiserson et al, 2013b, Kim et al, 2015, Ahmed et al, 2019, Baali et al, 2020]. Such a central role in driver gene and module identification has led to the design of many different approaches for defining and computing mutual exclusivity.…”

“…Such a central role in driver gene and module identification has led to the design of many different approaches for defining and computing mutual exclusivity. Some of these approaches are based on combinatorial definitions of mutual exclusivity [Vandin et al, 2012, Leiserson et al, 2013a, Basso et al, 2019, Ahmed et al, 2019, Song et al, 2020, Baali et al, 2020]. In most cases the combinatorial definitions are incorporated and tested within a driver gene or module identification framework, rather than as stand-alone ME tests.…”

“…We propose a network-centric framework to evaluate the pairwise significances found by statistical ME tests. It is important to make a distinction between the network-centric view of the current study and that of the previous studies employing both network data and the concept of ME [Ciriello et al, 2012, Leiserson et al, 2013b, Kim et al, 2015, Ahmed et al, 2019, Baali et al, 2020]. The latter are network-centric in the sense that the proposed ME tests are applied on interacting pairs or subnetworks as part of a more general goal of identifying cancer driver genes/modules.…”

A Network-Centric Framework for the Evaluation of Mutual Exclusivity Tests on Cancer Drivers

“…A phenomenon observed frequently in the data pertaining to the alterations that the tumors acquire is mutual exclusivity (ME); a driver mutation is less likely to occur in case of an earlier mutation that has common functionality in the same molecular pathway (Thomas et al, 2007;Yeang et al, 2008;Leiserson et al, 2016;van de Haar et al, 2019). Therefore several driver gene or module identification approaches employ ME detection as part of their problem definitions and optimization goals (Babur et al, 2015;Ciriello et al, 2012;Leiserson et al, 2013;Kim et al, 2015;Ahmed et al, 2019;Baali et al, 2020). Such a central role in driver gene and module identification has led to the design of many different approaches for defining and computing mutual exclusivity.…”

“…Such a central role in driver gene and module identification has led to the design of many different approaches for defining and computing mutual exclusivity. Some of these approaches are based on combinatorial definitions of mutual exclusivity (Vandin et al, 2012;Leiserson et al, 2013;Sarto Basso et al, 2019;Ahmed et al, 2019;Song et al, 2020;Baali et al, 2020). In most cases the combinatorial definitions are incorporated and tested within a driver gene or module identification framework, rather than as stand-alone ME tests.…”