SFCscore(RF): a random forest-based scoring function for improved affinity prediction of protein-ligand complexes.

A major shortcoming of empirical scoring functions for protein-ligand complexes is the low degree of correlation between predicted and experimental binding affinities, as frequently observed not only for large and diverse data sets but also for SAR series of individual targets. Improvements can be envisaged by developing new descriptors, employing larger training sets of higher quality, and resorting to more sophisticated regression methods. Herein, we describe the use of SFCscore descriptors to develop an improved scoring function by means of a PDBbind training set of 1005 complexes in combination with random forest for regression. This provided SFCscore(RF) as a new scoring function with significantly improved performance on the PDBbind and CSAR-NRC HiQ benchmarks in comparison to previously developed SFCscore functions. A leave-cluster-out cross-validation and performance in the CSAR 2012 scoring exercise point out remaining limitations but also directions for further improvements of SFCscore(RF) and empirical scoring functions in general.

MycPermCheck: the Mycobacterium tuberculosis permeability prediction tool for small molecules.

MOTIVATION: With >8 million new cases in 2010, particularly documented in developing countries, tuberculosis (TB) is still a highly present pandemic and often terminal. This is also due to the emergence of antibiotic-resistant strains (MDR-TB and XDR-TB) of the primary causative TB agent Mycobacterium tuberculosis (MTB). Efforts to develop new effective drugs against MTB are restrained by the unique and largely impermeable composition of the mycobacterial cell wall. RESULTS: Based on a database of antimycobacterial substances (CDD TB), 3815 compounds were classified as active and thus permeable. A data mining approach was conducted to gather the physico-chemical similarities of these substances and delimit them from a generic dataset of drug-like molecules. On the basis of the differences in these datasets, a regression model was generated and implemented into the online tool MycPermCheck to predict the permeability probability of small organic compounds. DISCUSSION: Given the current lack of precise molecular criteria determining mycobacterial permeability, MycPermCheck represents an unprecedented prediction tool intended to support antimycobacterial drug discovery. It follows a novel knowledge-driven approach to estimate the permeability probability of small organic compounds. As such, MycPermCheck can be used intuitively as an additional selection criterion for potential new inhibitors against MTB. Based on the validation results, its performance is expected to be of high practical value for virtual screening purposes. AVAILABILITY: The online tool is freely accessible under the URL http://www.mycpermcheck.aksotriffer.pharmazie.uni-wuerzburg.de