TY - JOUR TI - A statistical framework for assessing pharmacological responses and biomarkers using uncertainty estimates AU - Wang, Dennis AU - Hensman, James AU - Kutkaite, Ginte AU - Toh, Tzen S AU - Galhoz, Ana AU - GDSC Screening Team AU - Dry, Jonathan R AU - Saez-Rodriguez, Julio AU - Garnett, Mathew J AU - Menden, Michael P AU - Dondelinger, Frank A2 - Lehár, Joseph A2 - Weigel, Detlef A2 - Blocker, Alexander W VL - 9 PY - 2020 DA - 2020/12/04 SP - e60352 C1 - eLife 2020;9:e60352 DO - 10.7554/eLife.60352 UR - https://doi.org/10.7554/eLife.60352 AB - High-throughput testing of drugs across molecular-characterised cell lines can identify candidate treatments and discover biomarkers. However, the cells’ response to a drug is typically quantified by a summary statistic from a best-fit dose-response curve, whilst neglecting the uncertainty of the curve fit and the potential variability in the raw readouts. Here, we model the experimental variance using Gaussian Processes, and subsequently, leverage uncertainty estimates to identify associated biomarkers with a new Bayesian framework. Applied to in vitro screening data on 265 compounds across 1074 cancer cell lines, our models identified 24 clinically established drug-response biomarkers, and provided evidence for six novel biomarkers by accounting for association with low uncertainty. We validated our uncertainty estimates with an additional drug screen of 26 drugs, 10 cell lines with 8 to 9 replicates. Our method is applicable to any dose-response data without replicates, and improves biomarker discovery for precision medicine. KW - pharmacogenomics KW - biomarkers KW - machine learning KW - drug prediction KW - statistical inference KW - uncertainty estimation JF - eLife SN - 2050-084X PB - eLife Sciences Publications, Ltd ER -