DOLINA--docking based on a local induced-fit algorithm: application toward small-molecule binding to nuclear...

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Traditional docking methods often use rigid protocols, limiting accuracy.
• Protein flexibility, particularly local induced-fit, is crucial for understanding ligand binding.
• Nuclear receptors have buried binding sites where induced-fit is significant, especially for antagonists.

Purpose of the Study:

• To introduce and evaluate the Dolina docking software.
• To assess Dolina's ability to model protein flexibility and predict ligand poses accurately.
• To demonstrate Dolina's performance on nuclear receptor systems.

Main Methods:

• Dolina employs pharmacophore matching for pose generation and models induced-fit via side-chain rearrangement.
• Ligand flexibility is handled by screening a pool of low-energy conformers.
• Computational efficiency is achieved through residue grouping and side-chain conformer clustering.

Main Results:

• Dolina was tested on three nuclear receptors using 18 crystal structures.
• Accurate ligand poses (within 2.0 Å RMSD) were generated in 91% of cases.
• The correct pose was ranked first in 28% and within the top five in 76% of docking experiments.

Conclusions:

• Dolina effectively models local induced-fit in protein binding sites.
• The software demonstrates high accuracy in predicting ligand poses for nuclear receptors.
• Dolina is a valuable, freely available tool for academic research in drug discovery.