TY - GEN AB - A major challenge in precision oncology is accurately modeling the molecular machinery that governs cellular sensitivity or resistance to a given drug or drug combination. While predicting drug response is an active area, model performance and interpretability vary depending on the method and prediction task, and leaves room for improvement. In this talk, I'll discuss how these methods may be combined in a biology-centric framework that provides robust drug-response prediction, an interpretable latent space, and uncertainty quantification. AD - Oregon Health and Science University AD - Oregon Health and Science University AD - Oregon Health and Science University AU - Evans, Nathaniel AU - Blucher, Aurora AU - McWeeney, Shannon DA - 2020 DO - 10.6083/08612p16p DO - DOI ID - 8404 KW - Deep Learning KW - Machine Learning KW - precision oncology KW - drug response KW - graphical models KW - graph convolutional networks KW - visible neural networks KW - uncertainty quantification KW - bayesian modeling L1 - https://digitalcollections.ohsu.edu/record/8404/files/ResearchWeek.2020.Evans.Nathaniel.pdf L2 - https://digitalcollections.ohsu.edu/record/8404/files/ResearchWeek.2020.Evans.Nathaniel.pdf L4 - https://digitalcollections.ohsu.edu/record/8404/files/ResearchWeek.2020.Evans.Nathaniel.pdf LA - eng LK - https://digitalcollections.ohsu.edu/record/8404/files/ResearchWeek.2020.Evans.Nathaniel.pdf N2 - A major challenge in precision oncology is accurately modeling the molecular machinery that governs cellular sensitivity or resistance to a given drug or drug combination. While predicting drug response is an active area, model performance and interpretability vary depending on the method and prediction task, and leaves room for improvement. In this talk, I'll discuss how these methods may be combined in a biology-centric framework that provides robust drug-response prediction, an interpretable latent space, and uncertainty quantification. PB - Oregon Health and Science University PY - 2020 T1 - Integrating domain knowledge, interpretable deep learning, and uncertainty quantification for computational prediction of cell response to drug combinations TI - Integrating domain knowledge, interpretable deep learning, and uncertainty quantification for computational prediction of cell response to drug combinations UR - https://digitalcollections.ohsu.edu/record/8404/files/ResearchWeek.2020.Evans.Nathaniel.pdf Y1 - 2020 ER -