Hello GLYCAM mailing list



I have been tasked with running MD simulations of a oligosaccharide consisting of a modified D-Rha4Ac 1-2 repeats. The problem is that modifications are needed to the 4Ac group generating a “non-standard” derivative which does not have parameters in the forcefield (glycam), on every repeat.



I thought that it should probably be possible to combining GAFF or AMBER (ff14SB) with GLYCAM as seems to be basically what is happening when generating a glycoprotein. Especially as  all atoms in the modified substituent are standard atoms and bonds (N,C,O,H).



Having searched through the previous threads here I see that people have combined GAFF and GLYCAM before so this should be possible. I am now hoping that some kind person would be wiling to share a workflow to achieve this rather than reinventing the wheel.



In my mind there were two solutions. Generating a repeating oligosaccharide using glycam web with a 4NAc derivative providing a structure with aligned coordinates. The PDB file could then be edited manually or in a GUI  to give the correct substituent. I could use antechamber to generate GAFF parameters for the substituent attached to one monomer and edit the GLYCAM generated PDB with GAFF atom types/names matching. I would however suspect that this would require a frcmod file, the interphase between “GAFF” and “GLYCAM” would not exist in either FF, and potentially some additional files as the prepi file containing charges 



I also generated a “glycopeptide” using GLYCAM GUI. I could not use the modified substituent though could use glutamine and then replace it with the intended substituent later. Interestingly, the generated PDB file provided the “peptide” and “carbohydrate" as separate residues ("amber TER” separated) in the PDB. This file seemed to load correctly without any additional files/modifications in leap, only sourcing ff14SB and glycam. This made me think that this would only require modification of the substituent part of the PDB, which would be nice if other modifications would be needed later, and since I need the modifications on every monomer.



I have no idea if any of these thoughts are reasonable and at this point I am only curious if it is possible to combine AMBER/GAFF with GLYCAM parameter in a “simple” (like the suggestions outlined above) manner. If anyone who has performed simulations using combinations of GAFF and GLYCAM would be willing to share a bit more detailed and concrete outline of the steps involved in order to successfully achieve something like this, either here or directly by email, I would be extremely grateful.



Best regards

// Gustaf