.png){kind=icon}
返回
Building Block-Based Binding Predictions for DNA-Encoded Libraries
.png){kind=icon}
OA
AI
DOI：10.1021/acs.jcim.3c00588.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
DNA-encoded libraries (DELs) provide the means to makeand screenmillions of diverse compounds against a target of interest in a singleexperiment. However, despite producing large volumes of binding dataat a relatively low cost, the DEL selection process is susceptibleto noise, necessitating computational follow-up to increase signal-to-noiseratios. In this work, we present a set of informatics tools to employdata from prior DEL screen(s) to gain information about which buildingblocks are most likely to be productive when designing new DELs forthe same target. We demonstrate that similar building blocks havesimilar probabilities of forming compounds that bind. We then builda model from the inference that the combined behavior of individualbuilding blocks is predictive of whether an overall compound binds.We illustrate our approach on a set of three-cycle OpenDEL librariesscreened against soluble epoxide hydrolase (sEH) and report performanceof more than an order of magnitude greater than random guessing ona holdout set, demonstrating that our model can serve as a baselinefor comparison against other machine learning models on DEL data.Lastly, we provide a discussion on how we believe this informaticsworkflow could be applied to benefit researchers in their specificDEL campaigns.
Keyword:
CHEMICAL LIBRARIES
TECHNOLOGY
MOLECULES
DOCKING
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
5.3
论文数:
9.1K
被引数:
4.0W
机构
引用论文
Application of belief theory to similarity data fusion for use in analog searching and lead hopping
.png){kind=icon}
PREAI
分享收藏分享收藏分享收藏
查看更多内容