Shape-Based Descriptors for Efficient Structure-Based Fragment Growing - New publication in the Journal of Chemical Information and Modeling
17. November 2020
Fragment growing is an important paradigm in computer-aided drug design. To improve the speed and interactivity of fragment growing workflows, which are often built on old or general purpose tools, we have developed and recently published a novel shape-based algorithm. Using all the information available in a fragment growing scenario we generate cylindrical grids, which can be quickly compared on a numerical or bitwise level. We demonstrate that these shape descriptors can be used to perform fast and accurate fragment growings using test cases mined from public structural data. The fragment growing validation dataset can also be found here.