From: Tristan Croll (
Date: Sat Jun 21 2014 - 19:13:31 CDT

This is a very minor problem since I suspect anyone working with CHARMM-36 glycans will usually be using their own psfgen scripts anyway, but the heuristic that AutoPSF uses to determine chain type (protein/nucleic/other) gets mixed up when it encounters N-acetylglucosamine residues (and, I suspect, other N-acetylated sugars). It designates them as protein and attempts to apply N- and C-terminal patches accordingly. I suspect the problem here is that the N-acetyl amide has atoms named N, HN, C and O, which get mistaken for backbone atoms.

On a related note, may I ask if glycans are likely to be put on the "to-do" list to be automatically handled in a future release of VMD?