Benchmarking applicability of medium-resolution cryo-EM protein structures for structure-based drug design.

Area of Science:

• Structural Biology
• Computational Chemistry
• Drug Discovery

Background:

• Cryo-electron microscopy (cryo-EM) is increasingly used for protein structure determination.
• Many cryo-EM structures have resolutions between 3-5 Å, limiting their utility in in silico drug design.
• Accurate protein structures are crucial for effective in silico drug design and lead optimization.

Purpose of the Study:

• To evaluate the utility of medium-resolution cryo-EM structures for in silico drug design.
• To identify factors limiting ligand docking accuracy with cryo-EM data.
• To assess the impact of protein flexibility on docking performance.

Main Methods:

• Cross-docking simulations using Autodock-Vina.
• Comparison of docking success rates between cryo-EM and crystal structures.
• Analysis of resolution-dependent and independent factors affecting docking accuracy.
• Evaluation of flexibility implementation in docking tools.

Main Results:

• Only 20% of ligand docking attempts succeeded with 3-5 Å cryo-EM structures, compared to double the success rate with <2 Å crystal structures.
• Protein side-chain and backbone conformational heterogeneity was the primary resolution-dependent factor hindering docking.
• Intrinsic receptor flexibility contributed as a resolution-independent factor.
• Current docking tools with flexibility implementation improved success rates by only 10%, often limited by structural errors.

Conclusions:

• Medium-resolution cryo-EM structures present significant challenges for in silico drug design due to conformational heterogeneity.
• Existing ligand docking methods require enhancement to effectively utilize cryo-EM data.
• Further development in both cryo-EM modeling and docking algorithms is essential for advancing structure-based drug discovery.