One of the challenges in the small-molecule drug discovery field is to find therapeutic compounds that bind to targeted biomacromolecules with high affinity. Different computational methods being developed for small-molecule drug design need to be regularly tested on blind data sets to prevent researchers from exaggerating the performance of their methods by optimizing or overfitting their algorithms to reproduce known data sets. To this purpose, the Grand Challenge organized by the Drug Design Data Resource (D3R) community was conceived to provide a platform to evaluate methods in a blind way and to compare methods from different research groups. We present here our participation in D3R Grand Challenge 4, where we performed ligand-similarity based pose prediction and affinity rankings on the BACE-1 dataset. We used a protocol involving docking query ligands using binding pose information from the available co-crystal PDB structure of the ligand with the highest molecular similarity. Additionally, we submitted affinity rankings for 154 protein-ligand complexes using molecular mechanics combined with generalized Born surface area method (MM-GBSA). Our docking protocol based on HYBRID (OpenEye Scientific Software) was able to identify poses close to the native binding mode with an accuracy of &lt; 2 Å for most of the ligands. However, the affinity ranking results based on MM-GBSA calculations have poor correlation with experimental values with Spearman’s coefficients around 0.2. We also provide a critical evaluation of our predictions and possible ways to improve our docking/scoring workflow.

D3R Grand Challenge 4: Pose prediction and affinity ranking for BACE-1 inhibitors using ligand similarity and MM-GBSA calculations

2019 Fall National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.