Molecular fragments are an established part of drug design in both an experimental and a computational context. The very structure driven fragment-based drug design process creates many challenges for computational structure-based methods. Until now many of these challenges, such as structure-based fragment growing, were met with traditional software approaches, for example by extending a ligand and then completely re-docking it, or exploring the orientations of added fragments with brute force torsional tests.
A new generation of tools is necessary to specifically address fragments in their own right. In this BioSolveIT webinar Patrick will present a fragment growing tool using a novel, fast, and accurate shape-based algorithm. The tool also features an integrated interaction search, as well as a robust pose scoring function with gradient-based optimization.
The tool is validated on real fragment growings/replacements, automatically mined from the PDB that correspond to chemical series found in literature. This novel tool shall enable interactively explorative growing and put control over the growing process firmly back into the user’s hand.