UBC Theses and Dissertations
Deep learning-based approaches for predicting gene-regulating effects of small molecules Woo, Godwin Kar Min
Motivation: Recent advances in the areas of bioinformatics and predictive chemogenomics are poised to accelerate the discovery of small-molecule modulators of cellular processes. In that regard, combining large genomics and molecular data sources with recently emerged powerful deep learning techniques has the potential to revolutionize predictive biology. In this study, we present Deep Compound Profiler (DeepCOP), a deep learning (DL) based approach that can predict gene regulating effects of chemicals. This model, among many other potential applications, can be used for direct identification of a drug candidate causing a desired gene expression response, without utilizing any information on its interactions with particular protein target(s). Results: In this study we successfully combined molecular fingerprint descriptors and gene descriptors (derived from GO terms) to train deep neural networks (DNNs) that predict differential gene regulation endpoints collected in the Library of Integrated Network-Based Cellular Signatures (LINCS) database. The developed models yielded 10-fold cross validation AUC values of and above 0.80, as well as enrichment factors of >5. We validated the developed models using an external RNA-Seq dataset generated in-house that described the effect of three potent antiandrogens (with different modes of action) on gene expression in LNCaP prostate cancer cell line. The results of this pilot study demonstrate that DL models can effectively synergize molecular and genomic descriptors and can be used to screen for novel drug candidates with the desired effect on gene expression. We anticipate that such models can find a broad use in developing novel cancer therapeutics and can facilitate precision oncology efforts.
Item Citations and Data
Attribution-NonCommercial-NoDerivatives 4.0 International