Deep active learning-based experimental design to uncover synergistic genetic interactions for host-targeted therapeutics.

Motivation: High-throughput methods have advanced the study of host-virus interactions, but testing interactions between host gene pairs during infection remains labor intensive. Identification of multiple gene knockdowns that inhibit viral replication requires exploring a vast combinatorial space and is infeasible via brute-force experiments. Although active learning methods for sequential experimental design have shown promise, existing approaches have generally been restricted to single-gene knockdowns or small-scale double knockdown datasets. Results: Here, we present an integrated deep active learning (DeepAL) framework that incorporates information from a biological knowledge graph (SPOKE, the Scalable Precision Medicine Open Knowledge Engine) to efficiently search the configuration space of a large dataset of pairwise knockdowns of 356 human genes in HIV infection. Through representation learning, the framework is able to generate task-specific representations of genes while also balancing the exploration-exploitation trade-off to pinpoint highly effective double-knockdown pairs. In addition, we present an ensemble method for improved performance and an interpretation of the gene pairs selected by our algorithm through pathway analysis. To our knowledge, this is the first work to show promising results on double-gene knockdown experimental data of appreciable scale (356 by 356 matrix). Availability and implementation: https://github.com/LLNL/DeepAL.