bioRxiv · 10.64898/2026.03.31.715531
SHOT-CCR: Biologically guided adversarial training for test-time adaptation in cellular morphology
Abstract
AO_SCPLOWBSTRACTC_SCPLOWPervasive batch effects are a common issue, especially in recent large-scale Cell Painting datasets, which have been produced to aid AI-enhanced drug discovery efforts. Technical differences arising from experiments carried out in different batches can cause models to fail to generalize to unseen batches, despite good predictive performance "within batch". We propose a biologically grounded test-time adaptation framework, SHOT-CCR, which uses cell-invariant gradient reversal to decouple morphological signal from experimental confounders. Our approach performs 4.5% better than the current RxRx1 benchmark, classifying 1,139 classes of siRNA genetic perturbations with 91.6% accuracy. We deliver consistent results over four distinct cell types and two prominent Cell Painting datasets - RxRx1 and a subset of JUMP-CP. Across 484 classes of CRISPR perturbations in JUMP-CP our method improves accuracy by 15.7%.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Dee, W., Wenteler, A., Seal, S., Morris, O., Slabaugh, G.. 2026-04-02. SHOT-CCR: Biologically guided adversarial training for test-time adaptation in cellular morphology. https://doi.org/10.64898/2026.03.31.715531
Cite the original work for its findings. Save a collection to share your selection of sources.