Weird psfgen bug?

From: Tristan Croll (tristan.croll_at_qut.edu.au)
Date: Fri Oct 30 2015 - 03:22:06 CDT

Hi all,

I'm attaching a stream file, input PDB and the resulting PSF/PDB (to replicate it exactly, you'll need to load top_all36_na.rtf and top_all36_cgenff.rtf ahead of the streamfile, and apply 5TER and 3TER patches). What's got me confused is that a few of the hydrogens in the autopsf PDB file are missing their element assignments - e.g.

ATOM 25 HM21 2MA A 1 2.137 -2.186 -5.080 1.00 0.00 AO1 H
ATOM 26 HM22 2MA A 1 0.873 -2.410 -6.313 1.00 0.00 AO1 H
ATOM 27 HM23 2MA A 1 1.953 -1.088 -6.375 0.00 0.00 AO1
ATOM 28 C2' 2MA A 1 1.745 0.854 -0.366 1.00 0.00 AO1 C

.. despite being of the same atom type as others in the structure. In most cases this of course won't cause problems, but it plays merry hell with Gaussian since fftk writes them as element X in the .gau files. Easily fixed once you know what's going on, but a pain.

Cheers,

Tristan


This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:22:12 CST