P01-20

Induced-Fit Posing (IFP): A new pose prediction tool for hit to lead stage of drug discovery

Samuel TOBA *, Chiharu KONDA

OpenEye, Cadence Molecular Sciences
( * E-mail: samuel.toba@eyesopen.com )

Accurate prediction of binding poses is a cornerstone of structure-based drug design, essential for developing effective therapeutic agents. The accuracy of these predictions determines the efficiency of identifying and optimizing lead compounds, ultimately impacting the success rate of drug discovery projects. Accurate binding pose prediction is particularly achievable during the lead optimization phase, where the molecules of interest often share significant structural similarities with known crystallographic ligands. This similarity simplifies the docking process, allowing for more reliable predictions. However, the hit-to-lead stage introduces complexities that can challenge the reliability of these predictions due to structural diversity.
The introduction of Induced-Fit Posing (IFP) addresses these challenges by incorporating flexibility into the docking process, allowing for more accurate predictions of how diverse ligands bind to target proteins. Through a multi-step process involving pruning, hypothesis generation, MD simulation, and consensus scoring, IFP significantly enhances pose prediction accuracy. Retrospective cross-docking studies validate its effectiveness, showing a substantial improvement in successful predictions. By adopting IFP, researchers can better navigate the complexities of the hit-to-lead phase, predict compounds with diverse chemotypes, incorporate protein binding site flexibilities into their models, and advance their drug discovery efforts with greater confidence.