Neural-Network Scoring Functions Identify Structurally Novel Estrogen-Receptor Ligands

详细信息    查看全文

• 作者：Jacob D. Durrant ; Kathryn E. Carlson ; Teresa A. Martin ; Tavina L. Offutt ; Christopher G. Mayne ; John A. Katzenellenbogen ; Rommie E. Amaro
• 刊名：Journal of Chemical Information and Modeling
• 出版年：2015
• 出版时间：September 28, 2015
• 年：2015
• 卷：55
• 期：9
• 页码：1953-1961
• 全文大小：540K
• ISSN：1549-960X

文摘

The magnitude of the investment required to bring a drug to the market hinders medical progress, requiring hundreds of millions of dollars and years of research and development. Any innovation that improves the efficiency of the drug-discovery process has the potential to accelerate the delivery of new treatments to countless patients in need. 鈥淰irtual screening,鈥?wherein molecules are first tested in silico in order to prioritize compounds for subsequent experimental testing, is one such innovation. Although the traditional scoring functions used in virtual screens have proven useful, improved accuracy requires novel approaches. In the current work, we use the estrogen receptor to demonstrate that neural networks are adept at identifying structurally novel small molecules that bind to a selected drug target, ultimately allowing experimentalists to test fewer compounds in the earliest stages of lead identification while obtaining higher hit rates. We describe 39 novel estrogen-receptor ligands identified in silico with experimentally determined K_i values ranging from 460 nM to 20 渭M, presented here for the first time.