Electronic structure theory on modeling short-range noncovalent interactions between amino acids

Area of Science:

• Computational Chemistry
• Biophysics
• Molecular Interactions

Background:

• Short-range noncovalent interactions (NCIs) are crucial in chemical and biological systems.
• Accurately calculating these interactions, especially within the van der Waals envelope, presents a significant computational challenge.
• Existing computational methods often struggle with the precision required for these complex bindings.

Purpose of the Study:

• To introduce SNCIAA, a comprehensive database of benchmark interaction energies for short-range NCIs between amino acids.
• To systematically assess the performance of various computational methods in describing these interactions.
• To provide a reliable resource for the development and validation of future computational models.

Main Methods:

• Creation of the SNCIAA database using high-level coupled-cluster with singles, doubles, and perturbative triples/complete basis set [CCSD(T)/CBS] calculations.
• Inclusion of 723 benchmark interaction energies for neutral/charged amino acid dimers derived from protein X-ray crystal structures.
• Systematic evaluation of methods including MP2, DFT, SAPT, composite methods, semiempirical approaches, and IPML against the benchmark data.

Main Results:

• Dispersion corrections are essential for accurately describing NCIs, even in systems dominated by electrostatics like hydrogen bonds and salt bridges.
• MP2, ωB97M-V, and B3LYP+D4 demonstrated the highest reliability for short-range NCIs.
• SAPT requires the inclusion of the δMP2 correction for accurate short-range NCI description; IPML performance degrades at short ranges.

Conclusions:

• The SNCIAA database provides a crucial resource for advancing computational methods for noncovalent interactions.
• Accurate modeling of short-range NCIs necessitates careful consideration of dispersion effects.
• The findings guide the selection and improvement of computational tools for predicting NCIs across all interaction ranges.