QSAR-Based Virtual Screening of Potential Anti-Candida Hit Compounds

Candida albicans is the most frequently isolated pathogen in women and its rapid ability to develop resistance to first- and second-line drugs has boosted the search for new antifungal agents.1 In this study, we report the development of binary and continuous QSAR models useful for virtual screening2 of novel anti-candida compounds. Initially, 22,926 chemical structures with MIC data for C. albicans were curated and standardized according to the protocol proposed by Fourches and coworkers.3 Curated data was them submitted to QSAR modeling and validation using Random Forest algorithm and molecular fingerprints (Morgan and FeatMorgan). As a result, we have obtained statistically predictive QSAR models, with CCR ranging between 0.90–0.91 (binary models) and Q_ext^2 values ranging between 0.77–0.81 (continuous models). Finally, we applied the best models for virtual screening of eMolecules library, which allowed us to select 20 virtual hits for prospective in vitro assays.