Synergy aware flow matching improves single cell perturbation prediction
Synergy-Aware Residual Flow Matching for Single-Cell Perturbation Prediction
Machine LearningArtificial Intelligence
Summary
Predicting how single cells respond to genetic or chemical changes is tricky because most effects add up simply, but the interesting part comes from interactions between changes. The authors created SynGeneFlow, a method that focuses on these interactions by modeling the leftovers after accounting for simple effects. This approach uses biological insights about gene interactions and adjusts for how these change depending on the perturbation. Tests show it can more accurately forecast cell responses in various scenarios.
What this means in practice
- •For drug development teams: Model combined drug effects more accurately by capturing interactive gene responses beyond simple additive models.
- •For cell engineering engineers: Improve predictions of cell behavior under multiple genetic edits by focusing on interaction-driven changes in gene expression.
Authors
Jingrui Wu, Wangyun Dan
Abstract
Predicting single-cell responses to genetic and chemical perturbations is important for understanding regulatory mechanisms and prioritizing therapeutic interventions. Existing methods have made progress by modeling perturbation-conditioned expression shifts with neural networks, graph priors, foundation-model representations, or distributional generative models. However, combinatorial perturbation prediction remains challenging because much of the response can be explained by additive single-perturbation effects, while the biologically informative signal often lies in a smaller non-additive residual induced by perturbation interactions. Directly generating the full post-perturbation expression profile may therefore allocate substantial capacity to the dominant additive component and underemphasize synergy-specific structure. To address this issue, we propose \textbf{SynGeneFlow}, a synergy-aware flow matching framework that anchors the source distribution at an additive prior and learns a residual flow toward the true perturbed state. SynGeneFlow further incorporates perturbation-conditioned gene interaction modeling, where a static gene-gene mask provides biological structure and a perturbation-dependent deformation bias dynamically modulates gene-gene coupling. To preserve population-level response structure, we align both endpoint distributions and biological residual distributions with MMD-based regularization. We evaluate SynGeneFlow on genetic and drug perturbation benchmarks under observed-single, held-out, and cross-modality settings. The results demonstrate that residualizing flow matching around additive perturbation effects provides a biologically grounded and robust framework for single-cell perturbation response prediction.