Improving Predictions for Novel and Proprietary Compounds
Sanji Bhal
Abstract
It can sometimes be challenging to get accurate physicochemical property predictions for novel compounds from 'out-of-the-box' software solutions. ACD/Labs has been proactive in offering System Training and Accuracy Extender with pKa, logP, and solubility predictors to help overcome this. These tools facilitate the use of experimental measurements of proprietary/novel compounds to train predictor algorithms - allowing you to expand the 'vendor database' to reflect the chemistry space being investigated in-house.
Download
PDF format (118 KB)MS Word format (263 KB ZIP file)
