Data associated with "CFM-GP: Unified Conditional Flow Matching to Learn Gene Perturbation Across Cell Types"

Understanding gene perturbation effects across diverse cellular contexts is a central challenge in functional genomics, with significant implications for therapeutic discovery and precision medicine. While single-cell technologies enable high-resolution measurement of transcriptional responses, collecting such data remains expensive and time-intensive, especially when repeated for each cell type. Existing computational methods attempt to predict these responses but typically require separate models per cell type, limiting scalability and generalization.CFM-GP (<b>C</b>onditional <b>F</b>low <b>M</b>atching for <b>G</b>ene <b>P</b>erturbation) is a deep learning framework that models perturbation as a continuous transformation between control and perturbed gene expression distributions, conditioned on cell type. A single model generalizes across all cell types, eliminating the need for cell type–specific training.