Solving the scalability issue in quantum-based refinement: Q|R#1.

Area of Science:

• Computational chemistry
• Structural biology
• Biophysics

Background:

• Quantum-chemical calculations are essential for accurate biomacromolecule refinement.
• The computational cost of these calculations scales exponentially with system size, limiting their application to large biomolecules.
• This limitation, termed Q|R#1 in the Q|R software, necessitates novel approaches for efficient computation.

Purpose of the Study:

• To develop and validate a novel fragmentation method for accurate quantum-chemical refinement of large biomacromolecules.
• To address the computational intractability of quantum-chemical calculations for systems with many atoms.
• To implement this approach within the Q|R software package.

Main Methods:

• A divide-and-conquer strategy was employed, fragmenting the atomic model into smaller, manageable pieces.
• Noncovalent interactions were analyzed to create an interaction graph, which was then partitioned into clusters using a graph-clustering algorithm.
• Residues interacting with each cluster were assigned to a buffer region, forming fragments. Gradients were computed for each fragment and combined to yield total gradients for quantum-based refinement.

Main Results:

• The interaction graph-based fragmentation approach successfully enabled quantum-chemical refinement of a large biomacromolecular system.
• The method effectively managed the computational cost associated with large atomic models.
• Validation was performed on an amyloid cross-β spine crystal structure (PDB entry 2oNA).

Conclusions:

• The developed fragmentation method offers a viable solution to the computational challenges in quantum-chemical refinement of biomacromolecules.
• This approach allows for accurate calculations on large systems previously considered intractable.
• The Q|R software, incorporating this method, can be used for precise biomolecular modeling and refinement.