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

## TL;DR

Lilace is a new Bayesian model that improves the accuracy of analyzing FACS-based DMS data by better estimating variant effects and reducing false discoveries.

## Contribution

Lilace introduces a novel Bayesian statistical method for variant effect estimation in FACS-based DMS data.

## Key findings

- Lilace improves false discovery rates in FACS-based DMS experiments.
- The method maintains high sensitivity when applied to real-world datasets like OCT1 and Kir2.1.
- Validation on simulated data confirms Lilace's robustness and performance.

## 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.

The online version contains supplementary material available at 10.1186/s13059-026-03934-1.