From: Francesco Pietra (chiendarret_at_gmail.com)
Date: Wed Dec 23 2020 - 08:55:08 CST

Sorry, VMD 1.9.1 should be read 1.9.3
---------- Forwarded message ---------
From: Francesco Pietra <chiendarret_at_gmail.com>
Date: Wed, Dec 23, 2020 at 3:52 PM
Subject: Fake angles created by AutoPSF
To: VMD Mailing List <vmd-l_at_ks.uiuc.edu>

Hello
I am facing a puzzle with AutoPSF (VMD 1.9.1 linux).

I have to closely related homotrimers, built from X-ray diffr data, with
several S-S bonds.

The homotrimer at low pH allowed building correct psf/pdb and MD, so that I
did not check the psf of the other homotrimer, at high pH.
With the latter, minimization of psf/pdf gave error for the subuinit of
segn AP1

UNABLE TO FIND ANGLE PARAMETERS FOR SM SM SM (ATOMS 380 2044 380)
ATOM 380 SG CYS A 94 -9.311 -18.367 56.731 1.00 0.00 AP1

The reason is that the 380-2044 bond was created twice
psf BONDS
380 2044 380 2044 380 377 381 375

and the fake angle 380-2044-380 was also created
psf ANGLES
380 2044 2041 380 2044 380 380 377 379

These anomalies are repeated in the other two subunits.

I examined the starting pdb finding no anomalies at the various CYS
residues. I also checked the applied DISU, which looked like to be correct

Thanks for advice on where to search the error

francesco pietra