[Feature-dev] [ feature-Bugs-1313 ] ATOM-TYPE-IS-OTHER is non-zero even when all atoms being looked at have standard PDB naming

[noreply at simtk.org]

Bugs item #1313, was opened at 2010-08-18 04:51
You can respond by visiting: 
https://simtk.org/tracker/?func=detail&atid=145&aid=1313&group_id=16

Category: Application - featurize
Group: None
Status: Open
Resolution: None
Priority: 4
Submitted By: Grace Tang (gwtang)
Assigned to: Nobody (None)
Summary: ATOM-TYPE-IS-OTHER is non-zero even when all atoms being looked at have standard PDB naming

Initial Comment:
FEATURE reads in Amber files (amber*.dat) that assign partial charges as well as condense the atom types.  However, the Histidine residue (HIS) is not associated with any Amber information because Amber does not name it HIS.  Instead, Amber names it HID, HIE, or HIP depending on the protonation state.  As a result, whenever FEATURE see a HIS residue, it goes through the Amber parameter file and finds no match.  This means whatever properties relying on the Amber file get default values.  ATOM-TYPE-IS-OTHER is the default for ATOM-TYPE-IS.  Additional properties affected should include CHARGE, POS-CHARGE, NEG-CHARGE and maybe others.  I believe the code separately handles the RESIDUE-NAME-IS property, so HIS still gets mapped to RESIDUE-NAME-IS-HIS.

The above explanation is incomplete as non-zero ATOM-TYPE-IS-OTHER is also observed in sites without Histidine.  However, I do believe that Histidine is the bulk culprit.

Please see the attached files for more detail.  The example provided has ATOM-TYPE-IS-OTHER problems with Histidine and something else.

----------------------------------------------------------------------

You can respond by visiting: 
https://simtk.org/tracker/?func=detail&atid=145&aid=1313&group_id=16