Creating topology for chimeric DNA/RNA strand

GROMACS version: 2024.1
GROMACS modification: No

Hello everyone,
Looking for guidance on how to simulate a maybe niche system. It is a protein-oligonucleotide interaction in solution, however the oligo in question is a chimeric strand, meaning some nucleotides are DNA bases and some are RNA bases. I have been able to simulate this system and get reasonable results when the oligo is entirely RNA, however when I change some bases to DNA I get the following error message:

Fatal error:
The residues in the chain DT1ā€“RC78 do not have a consistent type. The first
residue has type ā€˜DNAā€™, while residue RG21 is of type ā€˜RNAā€™. Either there is a
mistake in your chain, or it includes nonstandard residue names that have not
yet been added to the residuetypes.dat file in the GROMACS library directory.
If there are other molecules such as ligands, they should not have the same
chain ID as the adjacent protein chain since itā€™s a separate molecule.

Which makes sense in retrospect since its changing residue type halfway through. However, since we technically have the parameters for each possible nucleotide, I wanted to check if it is possible to ā€œmergeā€ the DNA and RNA force fields into one and if people are aware of a method/tool to do so. Alternatively, Iā€™m open to any suggestions from people on how to handle such a system. For reference, Iā€™m using the AMBER ff14sb and XOL3 and OL15 forcefields.

Iā€™d be grateful for any advice.

Still interested in this if people have answers. I have tried manually merging force fields but the charges become non-integer. Therefore, if people know how to correct the charges of custom force fields that would also be helpful.

So Iā€™ve tried briefly to see if I could get it to work, turns out this isnā€™t an issue with the forcefields, itā€™s something that comes with gromacs. The file ā€˜residuetypes.datā€™ is included in the gromacs installation and basically lists which residues can be linked into a single molecule. Youā€™d need to find and modify this file to get this to work.

Iā€™ve tried creating a conda environment where I install gromacs and edit the residuetypes.dat, and pdb2gmx seems to work if all the RNA residues in residuetypes.dat are listed as DNA instead of RNA, but Iā€™m not enough of an expert in the program to know whether or not this will cause issues later on during the simulations.

(Also, since this requires editing the files in the gromacs folder rather than just altering the forcefields, this would have to be done on some separate installation that wouldnā€™t affect any of your other projects)

Hi thanks for your answer. Yes, I was able to copy the force field and residuetypes.dat file to my working folder so it uses both of these custom versions when selecting the field as I create the topology. I edited the residuetypes.dat so that it reads everything as RNA. The charges are still slightly off by 0.0002 from an integer. If you manually rebalance on the topol_RNA_chain.itp file by that small charge then my simulation ran fine and the results made sense based on what I expect from my experimental data and what I know from this system.

1 Like

It is not a good practice to edit the residuetypes.dat. Instead, you should edit the input PDB files to match the force field.

@scinikhil The problem is that this doesnā€™t seem to be an issue with a mismatch between coordinate file and forcefield, gromacs uses residuetypes.dat independent of which forcefield is chosen, and it doesnā€™t allow for a chimeric strand where DNA and RNA exists within the same molecule unless residuetypes.dat is modified