‹ back to 2026-09-10

An operational perturbation proteomics-based virtual cell model

Nature Bioinformatics AI Rui Sun, Liujia Qian, Yongge Li, Tong Liu, Honghan Cheng, Xuedong Zhang, Xueya Zhou, Yuecheng Zhan, Guangmei Zhang, Zhengchao Luo, Kunpeng Ma, Chunlong Wu, Dongchen Ji, Zhangzhi Xue, Hongxue Meng, Yuhang Xiang, Dingwei Lei, Qianhe Zhou, Wenbin Hu, Yuhan Deng, Lingling Tan, Qi Xiao, Zhiwei Liu, Lei Zeng, Liqin Qian, Xuan Zheng, Qiong Hu, Nianzi Luo, Weinan E, Peijie Zhou, Han Wen, Yi Zhu, Tiannan Guo 2026-09-09

TL;DR - This Nature study introduces ProteinTalks, a virtual cell model built from temporal protein-abundance measurements in systematically perturbed breast cancer cell lines. It is designed as an operational tool for multiple drug-discovery tasks.

Uses perturbation proteomics to model cellular responses over time.
Draws on systematically perturbed breast cancer cell lines.
Focuses on protein-abundance dynamics rather than static measurements.
The provided summary does not specify the model architecture, benchmarks, or quantitative results.

view merged work →