Characterizing functional consequences of DNA copy number alterations in breast and ovarian tumors by spaceMap

Abstract

We propose a novel conditional graphical model - spaceMap - to construct gene regulatory networks from multiple types of high dimensional omic profiles. A motivating application is to characterize the perturbation of DNA copy number alterations (CNAs) on downstream protein levels in tumors. Through a penalized multivariate regression framework, spaceMap jointly models high dimensional protein levels as responses and high dimensional CNAs as predictors. In this setup, spaceMap infers an undirected network among proteins together with a directed network encoding how CNAs perturb the protein network. spaceMap can be applied to learn other types of regulatory relationships from high dimensional molecular profiles, especially those exhibiting hub structures. Simulation studies show spaceMap has greater power in detecting regulatory relationships over competing methods. Additionally, spaceMap includes a network analysis toolkit for biological interpretation of inferred networks. We applies spaceMap to the CNAs, gene expression and proteomics data sets from CPTAC-TCGA breast (n=77) and ovarian (n=174) cancer studies. Each cancer exhibits disruption of 'ion transmembrane transport' and 'regulation from RNA polymerase II promoter' by CNA events unique to each cancer. Moreover, using protein levels as a response yields a more functionally-enriched network than using RNA expressions in both cancer types. The network results also help to pinpoint crucial cancer genes and provide insights on the functional consequences of important CNA in breast and ovarian cancers. The R package spaceMap - including vignettes and documentation - is hosted on https://topherconley.github.io/spacemap.

Keywords: Conditional graphical models; Integrative genomics; Network analysis; Proteogenomics.

Figure 1:. spaceMap integrative analysis pipeline.

Predictors (e.g., CNA) and responses (e.g., protein abundance) data are inputs to the model fitting stage, where the model is tuned by cross validation and aggregated across 1000 bootstrap ensemble networks through the Boot.Vote procedure. The Boot.Vote network is input to the network analysis stage, where biological function is layered onto the network. Finally the network is visualized with Cytoscape.

Figure 2:. hub-net simulation:

Edge-detection performance summarized by MCC, power, and FDR, across 100 replicates with sample size N = 250. The overall performance is further decomposed into response subnetwork PROT-PROT and the predictor→response subnetwork CNA→PROT. spaceMap.CV, space and scggm are learned under CV.Vote and spaceMap.boot is learned under Boot.Vote. All tuning parameters are chosen by 10-fold CV.

Figure 3:. BCPLS application:

Three GO-enriched modules from spaceMap prot-net. Large circles denote proteins belonging to enriched GO terms: cell proliferation (blue), transcription from RNA polymerase II promoter (orange) and ion transmembrane transport (purple). Rectangles denote CNA-hubs.