Accurate variant effect estimation in FACS-based deep mutational scanning data with Lilace

Abstract Deep mutational scanning (DMS) coupled with fluorescence-activated cell sorting (FACS) provides a high-throughput method to link genetic variants with quantitative molecular phenotypes. Analysis of these experiments is challenging due to measurement variance and the multidimensional FACS readout. However, no statistical method has yet been developed to address these challenges. Here we present Lilace, a Bayesian statistical model to estimate variant effects with uncertainty quantification from FACS-based DMS experiments. We validate Lilace’s performance and robustness using simulated data and apply it to OCT1 and Kir2.1 DMS datasets, demonstrating an improved false discovery rate while largely maintaining sensitivity.