Average Information Content Maximization—A New Approach for Fingerprint Hybridization and Reduction

摘要: Fingerprints, bit representations of compound chemical structure, have been widely used in cheminformatics for many years. Although fingerprints with the highest resolution display satisfactory performance virtual screening campaigns, presence a relatively high number irrelevant bits introduces noise into data and makes their application more time-consuming. In this study, we present new method hybrid reduced fingerprint construction, Average Information Content Maximization algorithm (AIC-Max algorithm), which selects most informative from collection fingerprints. This methodology, applied to ligands five cognate serotonin receptors (5-HT2A, 5-HT2B, 5-HT2C, 5-HT5A, 5-HT6), proved that 100 selected four non-hashed reflect almost all structural information required successful silico discrimination test. A classification experiment indicated representation is able achieve even slightly better than state-of-the-art 10-times-longer significantly shorter time.