Is it possible to combine the GLYCAM forcefield for carbohydrates with AMBER/GAFF?
I am looking at simulating a polysaccharide consisting of repeating non-standard/bacterial carbohydrates. I need to use a non-standard "derivative", which I am unable to include using glycam.org or leap.
I suppose what I am asking is how one would proceed here, is it even possible and also preferably then reasonable, to piece together a PDB file using with atom names and types from different forcefields like when preparing a LPS/glycoprotein according to the manual/tutorials. Or, should this be treated like a “novel” molecule and parameterised using the antechamber route?
I am a bit torn here as something close to "the base carbohydrate" can be generated pretty much as a “standard” carbohydrate. This seems beneficial as I could use FF parameters intended for this purpose.
However, I can only add an Acetyl group at the position where I need it, I would need a modified N-acetyl group at a “non-standard” position, which I am unable to insert using GLYCAM builder/leap. Hence, I would need to insert this at the “non standard position” manually, which may be possible, and also exchange/replace the methyl group. My suspicion is that while moving the NAc group may work, the modification would break GLYCAM compatibility as the resulting derivative/substituent is not available using the builder or in the manual?
My thought was if it was possible to maybe treat the derivative/substituent as a “protein” and just attach the sugar at the correct position. However, this would not be a standard O/N-link. Also, even if combining two different sets of parameters for the “substituent” part and the “sugar” part (gaff/amber and glycam) works, have I moved to far away from the glycam carbohydrate so that it would no longer be a reasonable representation and treating the polysaccharide as a “novel” structure would be more correct.
Asking a lot, obviously. Summarising, is it at all possible to combine FFs in a single structure using glycam and gaff/amber using non-standard linking. Would it be reasonable to combine parameters from different FFs in some manner (manual edits, connect info, different atom-types and FFs) or would treating the entire structure as a “novel” structure a better choice?
Best regards
// Gustaf
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue May 18 2021 - 06:30:02 PDT
