2020 年 2 月 25 日，北京生命科学研究所/清华大学生物医学交叉研究院黄牛实验室在《Frontiers in Pharmacology》杂志发表题为“Predicting or Pretending: Artificial Intelligence for Protein-Ligand Interactions Lack of Sufficiently Large and Unbiased Datasets”的文章，发现 AI 用于预测蛋白-配体相互作用时，常用的训练集（PDBbind和DUD-E)对AI模型训练存在严重数据偏向（data bias)，导致 AI 模型表现虚高，预测缺乏泛化能力和鲁棒性，误导本领域的方法发展和实际应用，并基于此提出了如何客观评估 AI 模型的观点和建议。

Qiuyu Fu made important contribution to this study.

Hong Lin, Xiaozhe Zhang, Li Liu, Qiuyu Fu, Chuanlong Zang, Yan Ding, Yang Su, Zhixue Xu, Sining He, Xiaoli Yang, Xiayun Wei, Haibin Mao, Yasong Cui, Yi Wei, Chuanzheng Zhou, Lilin Du, Niu Huang, Ning Zheng, Tao Wang, and Feng Rao

Abstract
FoxO1 is a conserved transcription factor involved in energy metabolism. It is tightly regulated by modifications on its mRNA and protein and responds to environmental nutrient signals. FoxO1 controls the transcription of downstream genes mediating metabolic regulation. Dysfunction of FoxO1 pathways results in several metabolic diseases, including diabetes, obesity, non-alcoholic fatty liver disease, and atherosclerosis. Here, we summarize the mechanism of FoxO1 regulation behind these diseases and FoxO1-related drug discoveries.

The ZGC Forum 2019 “ZGC International Frontier Science and Technology Achievements Exhibition” (中关村国际前沿科技成果展) was held from October 16th to 19th, 2019 in the ZGC Exhibition Center.
 
This exhibition focuses on the theme of “cutting-edge technology and future industries” and focuses on five areas: artificial intelligence and information technology, life sciences and biotechnology, intelligent manufacturing and high-end equipment, new materials and new energy, and cutting-edge science and technology. A total of 167 scientific and technological achievements from 116 units at home and abroad were displayed.
 
Dr. Yu Zhou and Nannan Hou participated in the exhibition on behalf of NIBS, and presented the project “Molecular Simulation and Innovative Drug Discovery”.