KORP-PL

KORP-PL implements a sidechain-free coarse-grained knowledge-based scoring function to evaluate protein–ligand interactions using a minimalist protein backbone and ligand-atom representation.

Key Features:

• Sidechain-Free Minimalist Representation: Represents proteins without sidechains, using a coarse-grained backbone and explicit ligand atoms to reduce model complexity.
• 3D Joint Probability Distribution Function: Employs a 3D joint probability distribution function as a knowledge-based statistical potential that encodes pairwise orientation and positional relationships between protein backbone and ligand atoms.
• High Performance in Docking and Screening: Demonstrated a twofold improvement in the median 5% enrichment factor on the DUD-E benchmark relative to Autodock Vina.
• Efficiency and Speed: Optimized for rapid scoring of large numbers of docking conformations to facilitate large-scale virtual screening.

Scientific Applications:

• Computational Drug Discovery: Used for virtual screening and prioritization of compound libraries by scoring protein–ligand docking conformations.
• Molecular Modeling: Applied to molecular modeling studies that require rapid evaluation of protein–ligand binding poses.

Methodology:

Uses a sidechain-free coarse-grained protein backbone representation and a 3D joint probability distribution function as a knowledge-based statistical potential over pairwise orientations and positions between protein backbone and ligand atoms to score docking conformations.