Non-integer total charge after partial charge replacement for a covalent Cys–maleimide patch

User discussions
1

Hi,

I’m building a protein–ligand system in GROMACS and I have a question about a non-integer total charge after applying a patch to a cysteine (thiol–maleimide addition), where the new covalent bond is formed between a ligand carbon (maleimide) and the cysteine sulfur (S).

I generated a “patch” with ACPYPE (AM1-BCC) for the reactive region (Cys + maleimide / part of the ligand). In ACPYPE, the full fragment has a total charge of 0, but when mapping it into the final system I only replace the charges of a subset of atoms (e.g., 14 atoms that exist in my system topology): on the Cys side S, CB, and 2 H, and on the ligand side 4 C, 2 O, 1 N, and 3 H(the maleimide ring). After replacing only that subset, the summed difference between the new charges and the original charges for those atoms is ~ −0.15 e, so the overall system charge becomes non-integer.

Since I can’t reparameterize the patch right now, I fixed this by distributing +0.15 only over a few atoms within the ligand (e.g., the imide N, 2 carbonyls and 2 carbons) so that the total system charge becomes an integer again. After that, grompp, genion, and energy minimization ran fine.

Questions:

  1. Is this approach (compensating ΔQ only within the ligand/patch atoms) acceptable / common practice?

  2. When only a small local region of a large molecule is chemically modified (e.g., a covalent conjugation) and therefore only a subset of atomic charges is replaced, is there any recommended procedure/best practice to handle the fact that the replaced charges will generally not preserve the same net charge as the original subset (i.e., to ensure consistency and an integer total charge) ?

Thanks!