Pock Drug: A Model for Predicting Pocket Druggability That Overcomes Pocket Estimation Uncertainties

Predicting:protein druggabilityis a key interest in the target identification phagre of chug discovery. Here, we assess the pocket estimation methods influence on drugg-ability predictions by comparing Statistical models constructed frorii pockets estimated using different pocket estimation methods: O proximity of either 4 or 5.5 angstrom to a cocrystallized ligancl or DoGSite and fpocket estimation methods. We developed Pock Drug) a tobliSt pocket druggability Modd that copes with uncertainties in pocket boundaries. It is based on a linear discriminant analysis from a, pool of 52 descriptors combined with a selection of the racist stable and efficient models using different pocket estimation methods. PockDrug retains the best combinations of three pocket properties which impact druggability geometry, hydrophobicity, and atomaticity. It results in an rage accuracy of 87.9% +/- 4.7% using a test set and exhibits higher accuracy (similar to 5-10%) than previous studies that used an identical apo set In conclusion, this study confirms the influence of pocket estimation on pocket druggability prediction and proposes PockDrug as a new model that nyereomes pocket estimation variability.