What are your thoughts on my MD step at preprocessing?

GROMACS version:2024.1
GROMACS modification: Yes/No
I merged heme and a ligand by manually editing gro and top instead of “-merge”, after which MD was performed. This result was unexpected for me because heme is not covalent binding to the protein, and it and the ligand are separated from the protein in less than 10ns of MD, but the ligand can only be injected once, so how do I make sure that the protein and heme are integrated as a whole, and MD of the ligand with the conjugated protein I constructed?

I don’t understand exactly what it is you are asking. Would you like the protein, heme and ligand to remain close together during the simulation, even if they are not covalently bound? If so, have a look at intermolecular_interactions in File formats - GROMACS 2024.1 documentation.