2026年9月9日，西湖大学郭天南等科学家在《自然》发表研究，开发了基于扰动蛋白质组学的虚拟细胞模型ProteinTalks。

人工智能赋能的虚拟细胞模型代表了计算机药物发现的一种新兴方法，但大多数现有方法缺乏大规模、时间分辨的扰动蛋白质组学数据以及用于预测治疗反应的可解释框架。在此，研究人员从系统扰动的乳腺癌细胞系中生成了超过3800万条时间分辨的蛋白质丰度测量数据，并开发了ProteinTalks这一虚拟细胞模型。ProteinTalks的核心在于这一大规模动态蛋白质组资源与模型架构的协同，从而实现了新的预训练框架，能够从时间蛋白质组轨迹中学习可迁移的动态潜在表征。通过对蛋白质如何条件性响应不同扰动进行建模，该方法使模型能够作为操作工具用于多种药物发现任务：预测药物疗效和协同作用、发现新药物组合、探究与耐药相关的蛋白质、对患者反应进行分层，以及为患者类器官优先筛选候选药物。它还表现出稳健的可迁移性，可扩展到细胞系之外的患者来源类器官和临床活检，并且在所评估的方案下通常比选定的基准实现取得更高性能。总之，ProteinTalks展示了可扩展地预训练可迁移动态表征如何使可操作、动态感知、基于蛋白质组学的虚拟细胞模型推动计算机药物发现。

附：英文原文

Title: An operational perturbation proteomics-based virtual cell model

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

Issue&Volume: 2026-09-09

Abstract: Artificial intelligence-empowered virtual cell models represent an emerging approach for in silico drug discovery1,2,3, yet most existing approaches lack large-scale, time-resolved perturbation proteomics data and interpretable frameworks for predicting therapeutic responses. Here we generated more than 38million temporal protein-abundance measurements from systematically perturbed breast cancer cell lines, and developed ProteinTalks, a virtual cell model. Central to ProteinTalks is the synergy of this large-scale dynamic proteomic resource and the model architecture, enabling a new pretraining framework that learns transferable dynamical latent representations from temporal proteome trajectories. By modelling how proteins respond conditionally to different perturbations, this approach enables the model to function as an operational tool for diverse drug discovery tasks: predicting drug efficacy and synergy, discovering new drug combinations, probing proteins associated with drug resistance, stratifying patient responses and prioritizing drug candidates for patient organoids. It also shows robust transferability, extending beyond cell lines to patient-derived organoids and clinical biopsies, generally achieving higher performance than the selected benchmark implementations under the evaluated protocols. Together, ProteinTalks shows how scalable pretraining of transferable dynamic representations enables operational, dynamics-aware, proteomics-based virtual cell models to advance in silico drug discovery.

DOI: 10.1038/s41586-026-11001-9

Source: https://www.nature.com/articles/s41586-026-11001-9