The lack of new drugs for Gram-negative pathogens is a global threat to modern medicine. The complexity of their cell envelope, with an additional outer membrane, hinders internal accumulation and thus, the access of molecules to targets. Our limited understanding of the molecular basis for compound influx and efflux from these pathogens is a major bottleneck for the discovery of effective antibacterial compounds. Here we analyse the correlation between the whole-cell compound accumulation of ~200 molecules and their predicted porin permeability coefficient (influx), using a recently developed scoring function. We found a strong linear relationship (75%) between the two, confirming porins key role in compound penetration. Further, the remarkable prediction ability of the scoring function demonstrates its potentiality to guide the optimization of hits to leads as well as the possibility of screening ultra-large virtual libraries. Eventually, the analysis of false positives, molecules with high-predicted influx but low accumulation, provides new hints on the molecular properties behind efflux.<br>
State of the Art Iterative Docking with Logistic Regression and Morgan Fingerprints
