DCAF: Unifying Data Validity, Evaluation Validity and Predictive Modeling for Single-cell Perturbation Simulation
- 1BGI Research, Wuhan
Abstract
Single-cell perturbation profiling is limited by cost and throughput, while computational methods remain constrained by inconsistent evaluation and underperformance relative to simple baselines. Here we present DCAF, a unified framework for single-cell perturbation simulation that that jointly addresses data validity, evaluation validity and predictive modeling. Among its three components, DCAF-data addresses the single-measurement limitation by reconstructing pre- and post-perturbation status using entropy-regularized optimal transport, and is demonstrated on both cell line and multi-cell type datasets. DCAF-criterion evaluates log fold changes in non-silenced genes to improve sensitivity to model error and robustness to zero inflation noise, validated on synthetic datasets. DCAF-model combines JEPA with explicit modeling for regulatory-dynamics modeling for interpretable prediction. Across benchmarks, DCAF outperforms state-of-the-art deep-learning methods, as well as mean and linear baselines in cross-gene, cross-cell-type and cross-region generalization. In adult Mouse brain, DCAF recapitulates known cell-type-specific regulatory heterogeneity of Dgcr8 knockdown at both cell type and gene levels. DCAF thus provides a rigorous and biologically grounded foundation for reliable perturbation simulation across development, disease and therapeutic discovery.
Related
- Perturbational Validity for Foundation Models of Brain Dynamics: A Controlled Proof-of-Principle Simulation
- Cause-excluded Effect Evaluation: a framework for the faithful assessment of genetic perturbation target identification
- Averaging in gene perturbation benchmarks hides perturbation-strength variation: Consequences for model evaluation
- Plate design confounds donor variation and model evaluation in single-cell perturbation screens
- Genetic Perturbation Prediction Models Fail Out-of-Distribution: A Rigorous Benchmark and Systematic Evaluation
BibTeX
@misc{xia2026dcaf,
title = {{DCAF: Unifying Data Validity, Evaluation Validity and Predictive Modeling for Single-cell Perturbation Simulation}},
author = {T. Xia and C. Lei and N. Chen and H. Hou and H. Zhang and H. Qin and D. Wang and Q. Li and Y. Zhang and S. Fang and Y. Li},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.13.758713},
url = {https://www.biorxiv.org/content/10.64898/2026.10.13.758713v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - DCAF: Unifying Data Validity, Evaluation Validity and Predictive Modeling for Single-cell Perturbation Simulation
AU - T. Xia
AU - C. Lei
AU - N. Chen
AU - H. Hou
AU - H. Zhang
AU - H. Qin
AU - D. Wang
AU - Q. Li
AU - Y. Zhang
AU - S. Fang
AU - Y. Li
PY - 2026
DA - 2026/10/13
PB - bioRxiv
DO - 10.64898/2026.10.13.758713
UR - https://www.biorxiv.org/content/10.64898/2026.10.13.758713v1
AB - Single-cell perturbation profiling is limited by cost and throughput, while computational methods remain constrained by inconsistent evaluation and underperformance relative to simple baselines. Here we present DCAF, a unified framework for single-cell perturbation simulation that that jointly addresses data validity, evaluation validity and predictive modeling. Among its three components, DCAF-data addresses the single-measurement limitation by reconstructing pre- and post-perturbation status using entropy-regularized optimal transport, and is demonstrated on both cell line and multi-cell type datasets. DCAF-criterion evaluates log fold changes in non-silenced genes to improve sensitivity to model error and robustness to zero inflation noise, validated on synthetic datasets. DCAF-model combines JEPA with explicit modeling for regulatory-dynamics modeling for interpretable prediction. Across benchmarks, DCAF outperforms state-of-the-art deep-learning methods, as well as mean and linear baselines in cross-gene, cross-cell-type and cross-region generalization. In adult Mouse brain, DCAF recapitulates known cell-type-specific regulatory heterogeneity of Dgcr8 knockdown at both cell type and gene levels. DCAF thus provides a rigorous and biologically grounded foundation for reliable perturbation simulation across development, disease and therapeutic discovery.
ER -