Disjunctive Bayesian Network Infers Cancer Progression Network
Amir Asiaee Taheri, PhD Ohio State university
Cancer is an evolutionary process that can be modeled as a sequence of genetic alterations throughout the tumor cell population. Each new driver alteration confers a selective growth advantage to the cell and sweeps through the population, which results in clonal expansion. But the order in which accumulating alterations occur is not arbitrary. Inferring the order of events leading to cancer has been shown to have diagnostic and prognostic importance but is a challenging problem due to the lack of longitudinal samples from tumors.
In this talk, I am going to discuss our novel scalable algorithm for inferring cancer progression networks. To model cancer progression, we introduce the Disjunctive Bayesian Network (DBN), which is a discrete Bayesian Network (BN) with a particular family of local conditional probability distributions. We then use a genetic algorithm to learn the structure of DBN from cross-sectional cancer data. We characterize an equivalence relation over DBNs and speed up our algorithm by restricting our search space to a single representative graph of each equivalence class. Finally, we show that the progression networks inferred by our method for colon, bladder, and skin cancers match the biological facts known about these cancers.
2525 WEA Suite 1100 Conf Room 11105A 14 February 2020 2:30pm