AMBER Archive (2000)

Subject: Re: 2 questions

From: Bill Ross (ross_at_cgl.ucsf.edu)
Date: Wed Nov 15 2000 - 21:48:17 CST

        (1) In running simple MD, I often get an error called "Trace/BPT Trap".
        Is there a likely reason for getting this error?

Could be a core dump - does anything appear in the .out file?
What hardware are you running on?
        
        (2) I am interested in doing the following simple MD simulation:
        load a pdb
        create the .top and .crd files needed to go on to sander
        run simple md to get a picture of flexibility of the eqm structure
        (the protein is made of 2 chains). I
        have fixed my cys and his groups and they seem to be read fine. When
        AMBER reads the pdb,

Presumably you are talking about leap here..

        it seems to put extra atoms on terminal groups; the
        extra atoms do not have a noticeable TYPE. My going into the pdb file
        created by AMBER (after it read the original pdb) and deleting these
        lines allows AMBER to read it and create the .top and .crd files.

I would think the H's get re-added.

        However, I feel that I am doing something wrong here. Am I supposed to
        use connect atoms or something?
        
You should read up on the types of terminal residues the amber
forcefields support. If you don't like the default type, you can
change what gets applied to your system either by copying the
default leaprc to your working directory and changing the pdb
residue map, or by using loadpdbusingsequence; the 1st is easier
and is ok as long as it's what you always want for the default
in that directory.

Bill Ross