UniPert-G2CP bridges genetic and chemical screens from molecular representation to phenotype modeling
9 affiliations
- 1Hunan Provincial Key Laboratory on Bioinformatics, School of Computer Science and Engineering, Central South University, Changsha 410083, China; AI for Life Sciences Lab, Tencent, Shenzhen, China
- 2Hunan Provincial Key Laboratory on Bioinformatics, School of Computer Science and Engineering, Central South University, Changsha 410083, China
- 3AI for Life Sciences Lab, Tencent, Shenzhen, China; Tsinghua-Peking Joint Center for Life Sciences, School of Life Sciences, Tsinghua University, Beijing 100084, China
- 4AI for Life Sciences Lab, Tencent, Shenzhen, China; Department of Computer Science, City University of Hong Kong, Hong Kong, China
- 5AI for Life Sciences Lab, Tencent, Shenzhen, China
- 6AI for Life Sciences Lab, Tencent, Shenzhen, China; Department of Computer Science, Hong Kong Baptist University, Hong Kong, China
- 7AI for Life Sciences Lab, Tencent, Shenzhen, China. Electronic address: [email protected]
- 8Hunan Provincial Key Laboratory on Bioinformatics, School of Computer Science and Engineering, Central South University, Changsha 410083, China. Electronic address: [email protected]
- 9AI for Life Sciences Lab, Tencent, Shenzhen, China. Electronic address: [email protected]
Abstract
Artificial Intelligence virtual cell (AIVC) holds transformative potential for biomedical research. Central to this vision is the systematic modeling of genetic and chemical perturbation phenotypes to accurately predict cellular dynamic states from diverse interventions. However, disparities in screening agents, library scales, experimental technologies, and data production efficiency hinder the integration, modeling, and analysis of the cross-data. Here we present UniPert-G2CP , a two-phase deep learning approach comprising i) UniPert, a multimodal molecular representation model that bridges genetic and chemical domains, and ii) G2CP ( Genetic-to-Chemical Perturbation transfer learning), which systematically transforms CRISPR screen-based genetic insights into chemical perturbation modeling for cost-effective in silico drug screening. UniPert not only encodes multimodal perturbagens into a unified functionally interpretable sematic embedding space, but also improves phenotypic effect prediction for previously unseen gene perturbations and drug treatments. Building upon UniPert, G2CP successfully modeled large-scale cellular post-perturbation states spanning 4,994 gene and 7,821 compound perturbagens, while reducing modeling data costs by over 60%. We demonstrate that UniPert-G2CP enables efficient, generalizable simulations of multicellular, multi-domain perturbation cause-effect spaces, revealing differential cellular biological causality and informing mechanism-driven therapy. UniPert-G2CP opens new avenues for biological causal foundation model building, AIVC creation, and AI-powered precision medicine.
Related
- Sliding Window Attention with Auditable Chemical Sinks for Long-Sequence Molecular Transformation Modeling
- Multimodal Representation Learning for Exploring Natural Product Chemical Space
- Count your bits: fingerprint benchmarking to assess broad chemical space representation
- Weighted Views Representation with Multi-Head Coulomb Attention for Molecular Property Prediction
- On the connection between the exact factorization and the Born–Huang representation of the molecular wavefunction
BibTeX
@article{li2026unipertg2cp,
title = {{UniPert-G2CP bridges genetic and chemical screens from molecular representation to phenotype modeling}},
author = {Yiming Li and Min De ZENG and Jun Zhu and Linjing Liu and Fang Wang and L L Huang and Fan Yang and Min Li and Jianhua Yao},
journal = {Cell},
year = {2026},
month = jul,
doi = {10.1016/j.cell.2026.06.005},
url = {https://doi.org/10.1016/j.cell.2026.06.005},
note = {CC BY},
}
RIS
TY - JOUR
TI - UniPert-G2CP bridges genetic and chemical screens from molecular representation to phenotype modeling
AU - Yiming Li
AU - Min De ZENG
AU - Jun Zhu
AU - Linjing Liu
AU - Fang Wang
AU - L L Huang
AU - Fan Yang
AU - Min Li
AU - Jianhua Yao
PY - 2026
DA - 2026/07/25
JO - Cell
DO - 10.1016/j.cell.2026.06.005
UR - https://doi.org/10.1016/j.cell.2026.06.005
AB - Artificial Intelligence virtual cell (AIVC) holds transformative potential for biomedical research. Central to this vision is the systematic modeling of genetic and chemical perturbation phenotypes to accurately predict cellular dynamic states from diverse interventions. However, disparities in screening agents, library scales, experimental technologies, and data production efficiency hinder the integration, modeling, and analysis of the cross-data. Here we present UniPert-G2CP , a two-phase deep learning approach comprising i) UniPert, a multimodal molecular representation model that bridges genetic and chemical domains, and ii) G2CP ( Genetic-to-Chemical Perturbation transfer learning), which systematically transforms CRISPR screen-based genetic insights into chemical perturbation modeling for cost-effective in silico drug screening. UniPert not only encodes multimodal perturbagens into a unified functionally interpretable sematic embedding space, but also improves phenotypic effect prediction for previously unseen gene perturbations and drug treatments. Building upon UniPert, G2CP successfully modeled large-scale cellular post-perturbation states spanning 4,994 gene and 7,821 compound perturbagens, while reducing modeling data costs by over 60%. We demonstrate that UniPert-G2CP enables efficient, generalizable simulations of multicellular, multi-domain perturbation cause-effect spaces, revealing differential cellular biological causality and informing mechanism-driven therapy. UniPert-G2CP opens new avenues for biological causal foundation model building, AIVC creation, and AI-powered precision medicine.
SN - 0092-8674
SN - 1097-4172
ER -