Support for biomolecular mutations (Protein/DNA/RNA base modifications) in free energy calculations

[Heezch]

Goal

My goal with OpenFE is to calculate binding free energies for DNA–peptide complexes and, eventually, to perform alchemical transformations of DNA bases and base modifications. This appears related to #682, where amino acid mutations were discussed, but it was noted that treating an alchemical peptide or protein using an appropriate biomolecular force field was not considered feasible at the time. Relatedly, in #889 the workaround proposed in #682 was applied, but convergence issues were reported. That approach relies on treating the peptide as a small molecule, which becomes a fundamental limitation when considering DNA base mutations or mutations at protein–protein or protein–DNA interfaces.

In the short term, I would like to be able to run absolute binding free energy (ABFE) calculations for DNA–peptide complexes.
In the longer term, I am interested in alchemical mutations of DNA bases and chemically modified bases.

Current limitations

• ABFE currently requires the disappearing species to be represented as a SmallMoleculeComponent.
• SmallMoleculeComponents are constrained to OpenFF SMIRNOFF-based parameterization.
• Canonical peptide and DNA residues are already well parameterized using established biomolecular force fields. Generating charges via AM1-BCC, Gasteiger, or similar small-molecule workflows does not seem appropriate for these systems.
• There is no current support for biomolecular absolute or relative binding free energy calculations involving protein, DNA, or RNA mutations.

Questions for the developers

• Is there a roadmap for supporting biomolecular mutations in OpenFE?
• Would a dedicated BiomolecularRBFE protocol be considered or accepted as a community contribution?
• Are there recommended workarounds in the interim, such as using pre-parameterized residues, hybrid approaches, or custom protocol extensions?

My position

• I am willing to invest development time if there is a viable and supported path forward.
• I am seeking guidance on whether this direction aligns with OpenFE’s design philosophy and long-term goals.