This is an excellent tool for quick and easy modelling of glycans using MD. However, I have noticed a bug in the newer carbohydrate builder:
When I build an arabinoxylan fragment,
the system adds an extra Laraf residue to one of the central Dxylp (the one closer to the non-reducing end), making it double substituted. This only appears in the "structure.pdb" and "unminimised-gas" files, not the solvated tip3p and tip5p ones. Interestingly, this does not occur when the terminal Dxylp has an alpha-OH.
So far I haven't noticed this with any other structures.
In the meantime I can use the legacy builder for this particular system and any others that cause problems.