Identifying key somatic alterations in cancer is a critical step in understanding the mechanisms that ultimately determine a patient's treatment outcome. High-throughput data are providing a comprehensive view of these molecular changes for individual samples, and new technologies allow for the simultaneous genome-wide assay of genome copy number variations, gene expression, DNA methylation, and epigenetics of patient tumor samples and established cancer cell lines. Analyses of current datasets find that genetic alterations between tumors can differ but often involve common pathways. It is therefore critical to identify relevant pathways involved in cancer progression and detect how they are altered in different patients. This work presents a novel method called PARADIGM for inferring tumor-specific genetic pathway activities incorporating curated gene-pathway interactions. A gene is modeled by a factor graph as a set of interconnected variables encoding the expression and known activity of itself and its upstream and downstream products, allowing the incorporation of many types of -omic data as evidence. The method predicts the degree to which a given pathway's activities (e.g. internal gene states, interactions, or high-level outputs) are altered in the tumor sample using probabilistic inference. Compared to a competing pathway activity inference approaches, PARADIGM identifies altered activities in cancer-related pathways with fewer false-positives, as shown in glioblastoma multiform (GBM), ovarian (OV) and breast cancer datasets. PARADIGM also identified consistent pathway-level activities for subsets of the GBM and ovarian serous cystadenocarcinoma patients that are overlooked when genes are considered in isolation. Furthermore, grouping GBM and OV patients based on their significant pathway perturbations divides them into clinically-relevant subgroups having significantly different survival outcomes. Further analysis was done using an integrated pathway termed the SuperPathway that gives a more consistent global view of biology, illustrated through the analysis of ovarian, breast, and cross-cancer analysis. Finally, PARADIGM was used to simulated knockdowns in a pathway model as an approach to understand drug effects and provide a rational approach towards combination therapies. These findings suggest that therapeutics might be personalized and chosen to hit target genes at critical points in the commonly perturbed cancer pathway(s) of specific patients using this model.

Background

Methods

Results

Conclusions

Impact