New Atomic Correction Boosts Protein Energy Predictions

Solvation free energy drives many biological processes. The usual way to break this energy down into parts for each atom often overestimates the true values when used with 3D‑RISM theory. Researchers created a fresh correction called MILC‑AD. Instead of relying on bulk volume, this approach uses atom‑by‑atom interaction energies as its guide. The goal is to give a more reliable, site‑specific view of how molecules behave in solution.

The team tested the new method on a large set of 628 small molecules from the FreeSolv database. They compared the results to benchmark BAR calculations, averaging ten different shapes for each molecule. The mean absolute deviation came out at 0.57 kcal/mol, which is a solid improvement over earlier attempts. This shows the correction works well across many chemical types.

To prove the concept on a real protein, the researchers applied MILC‑AD to the 36‑residue villin headpiece (HP36). The analysis highlighted where packing inside the protein, the solvation free energy, and entropy balance each contribute. It revealed how the hydrophobic core assembles cooperatively, with each site playing a distinct role.

Overall, MILC‑AD provides a practical tool for scientists who need detailed, quantitative thermodynamic data for large biomolecular systems. By delivering site‑resolved insights without the usual overestimation, the method opens new possibilities for drug design and protein engineering.