Increasing the Accuracy of Optipharm's Virtual Screening Predictions by Implementing Molecular Flexibility

Recently, a new piece of software called OptiPharm has been proposed to optimize the similarity between two given molecules. A comprehensive study proved it was very competitive compared with state-of-the-art algorithms such as WEGA and ROCS. However, all of these methods, including OptiPharm, assume the proteins as rigid, resulting in poor or inefficient predictions. The consideration of conformational changes and thus the molecule's flexibility is necessary. In this work, we have extended the OptiPharm's functionality by applying a methodology that considers the flexibility of the molecules. Apart from that, the new OptiPharm presents some strengths regarding its previous version. More precisely, it reduces the search space dimension and introduces circular limits for the angular variables to enhance searchability. As results will show, these improvements help OptiPharm to achieve better predictions.

Automated summary

OptiPharm is a new software that optimizes the similarity of molecules and competes well with other algorithms. However, all methods, including OptiPharm, assuming proteins as rigid can lead to inaccurate predictions. This study extends OptiPharm's functionality by considering molecular flexibility, introducing improvements that enhance prediction accuracy.