Control-Anchored Residual Flow Matching Conditioned on Gene Geometry for Virtual Cell Perturbation Modeling

2026-08-07

AI Summary

This study addresses the challenge of accurately predicting single-cell transcriptional responses to unseen genetic perturbations and drug combinations while avoiding the misattribution of stable gene co-expression patterns as perturbation-induced effects. To this end, the authors propose GeneGeoFlow, a novel method that incorporates perturbation-conditioned, multi-scale gene geometric structures as priors. It employs a perturbation-gated module to dynamically select relevant structural information and constructs a residual flow model anchored on control samples to disentangle stable gene associations from intervention-specific responses. Integrating multi-scale spectral coordinates, anchor-based residual flows, unpaired optimal transport training, and a Delta-correlation objective function, GeneGeoFlow achieves a Pearson Delta score of 0.8979 on the Norman additive benchmark and 0.9088 across five held-out drug combinations in ComboSciPlex, substantially outperforming existing approaches.

Abstract

A central task in virtual cell modeling is predicting single-cell transcriptional responses to unseen genetic perturbations and drug combinations, and biological networks provide valuable priors on gene relationships. Existing graph-based models commonly use the same network to structure gene representations and mediate intergene interactions, thereby implicitly treating stable associations as perturbation-response pathways. Gene Ontology and control-derived coexpression networks encode relatively stable relationships rather than intervention-specific response directions or magnitudes. We therefore propose GeneGeoFlow, which conditions a control-anchored residual flow on gene-wise geometry derived from biological networks to learn intervention-specific transcriptional responses. GeneGeoFlow derives multi-scale spectral coordinates from Gene Ontology and control-derived coexpression networks. A perturbation-conditioned, gene-wise gating module selects relevant structural scales and network sources, yielding intervention-specific gene geometry. The resulting geometry conditions a control-anchored residual flow without explicitly propagating target-derived signals along the graph. Condition-wise optimal transport couples unpaired control and perturbed populations for training, while a Delta-correlation objective aligns the predicted and observed condition-level expression-shift directions. GeneGeoFlow achieves Pearson Delta scores of 0.8979 on the Norman additive benchmark and 0.9088 on five held-out drug combinations in the fixed ComboSciPlex test split. These results support perturbation-conditioned gene geometry as an effective structural prior for intervention-specific response prediction, without conflating stable gene relationships with response propagation.

Problem

• virtual cell perturbation
• transcriptional response prediction
• gene geometry
• biological networks
• intervention-specific response

Innovation

• gene geometry
• residual flow matching
• perturbation-conditioned modeling
• multi-scale spectral coordinates
• control-anchored flow