Generalizable single-cell perturbation response prediction using energy-guided flow matching

Abstract

Predicting phenotypic and transcriptional responses to perturbations at single-cell resolution provides a powerful tool for probing biological systems. However, existing methods typically rely on fixed mappings learned during training, making it challenging to calibrate distribution shifts or adapt to novel perturbation conditions during inference. Here, we present scEGFlow, an energy-guided flow matching framework that dynamically bridges control and perturbed cellular states. scEGFlow models continuous transitions from control cell populations to perturbed states using conditional flow matching. It then applies condition-specific energy gradients to correct and steer these predictions, enabling flexible adjustments without retraining the flow model. Evaluations across benchmarks spanning imaging phenotypes and transcriptomic profiles show that scEGFlow outperforms existing methods in reconstructing response distributions under both seen and unseen perturbation conditions, faithfully preserving cellular manifold geometry and population heterogeneity. This advantage is notable when adapting to new conditions with only a few measured cells, consistently improving prediction accuracy. Furthermore, scEGFlow accurately recapitulates perturbation-induced up- and down-regulation patterns across consensus gene expression signatures, where energy guidance improves the agreement between predicted and observed regulatory directions. Ultimately, these findings demonstrate that scEGFlow provides a modular, generalizable, and steerable solution for single-cell perturbation modeling. By grounding generative flows in learned biological landscapes, this architecture establishes a new computational paradigm for navigating and manipulating cellular behavior in silico.

The authors’ abstract, as arXiv lists it. arXiv’s listings, abstracts included, are free for anyone to reuse.

More in Genomics

• arXiv preprint · 29 September 2026

  Stochastic gradient descent on the epigenetic landscape: a unified framework for cellular plasticity, tumor heterogeneity, and the asymptotic irrelevance of fitness

  Artur C. Fassoni
• arXiv preprint · 26 September 2026

  Structured Population Models for Follicular Development

  Edilbert Christhuraj, Claudio Iuliano and Alexander Lange
• arXiv preprint · 2 October 2026

  Multimodal reasoning for broadly neutralizing antibody discovery from label-free human B cell repertoires across virus families

  Hantao Lou, Jianqing Zheng, Can Yue and 21 others
• arXiv preprint · 22 September 2026

  Entropy-Driven Initiation and Cytoskeletal Viscoelasticity in Endocytosis: An Onsager Variational Framework

  Jinjie Liu, Zhongcan Ouyang and Hao Wu
• arXiv preprint · 7 October 2026

  Label-free cell counting and viability prediction with brightfield imaging and deep learning

  Amir Reza Vazifeh, Christian Zeigler, Sornanathan Meyyappan and 2 others

All new papers in Genomics