Skip to main content
Fig. 5 | Journal of Cheminformatics

Fig. 5

From: Dimensionally reduced machine learning model for predicting single component octanol–water partition coefficients

Fig. 5

Parity density plots of final testing data (N = 2,713) for the deployed MF-LOGP algorithm as well seven peer-reviewed methods that have demonstrated strong predictive capabilities, but all require structural information. MF-LOGP outperforms four of the current models and has a close performance to the EPA’s KOWWIN algorithm. a MF-LOGP, b XLOGP3, c MLOGP, d iLOGP, e ALOGPS, f KOWWIN, g WLOGP, h SILICOS-IT, i AAM

Back to article page