In Search of Predictive Models for Inhibitors of 5-Alpha Reductase 2 Based on the Integration of Bioactivity and Molecular Descriptors Data
January 2014
in “
IWBBIO
”
Studysummary This study used machine learning to develop classifiers for identifying effective inhibitors of 5α-reductase isozyme 2, achieving high performance in distinguishing potent from weak inhibitors. Our plain-language summary of this paper — not a Tressless recommendation.
The study focused on developing predictive models for inhibitors of 5-alpha reductase 2, an enzyme linked to conditions like baldness and prostate disorders. Using a dataset from the ChEMBL database, the researchers applied machine learning techniques, specifically random forests and support vector machines, to create classifiers for virtual screening. These models aimed to prioritize compounds for further investigation. The evaluation of the models showed that both algorithms performed similarly well, demonstrating high sensitivity, specificity, precision, F-score, and accuracy in distinguishing between potent and weak inhibitors. This research highlighted the potential for these models to aid in the discovery of more effective inhibitors with fewer side effects than existing drugs like finasteride and dutasteride.
