AMBER Archive (2004)

Subject: AMBER: using LJ 10-12 non bonded potential for glycoproteins in Amber8

From: Tom Turner (thomas.turner_at_imperial.ac.uk)
Date: Wed Mar 17 2004 - 12:49:18 CST

Dear Amberphiles,

I am modelling a protein with a number of sugar residues attached which
require me to use the GLYCAM paramater file "glycam_2000a.dat". This
parameter file makes use of HBOND terms that use the Leonard Jones 10-12
potential which (correct me if i'm wrong) has been phased out of the
standard Amber force field since 1994. Thus, sander complains that it
needs to be recompiled with the "-DHAS_10_12" flag to activate the LJ
potential.

So, this I have done by including the -DHAS_10_12 flag in the
AMBERBUILDFLAGS section of the config.h file. I can then successfully
compile amber8 but running "make test" produces an error during the
./Run.tleap test - "Program error in leap". This error did not occur
without the flag. Is this a bug?

So, on the reflector archive there exists this alternate suggestion for
dealing with GLYCAM's non-bonded parameters:

"....just remove HBOND information from your glycam parm file. "

i.e. strip out the HBON terms from glycam_2000a.dat then the -DHAS_10_12
flag won't be required. Is this a valid thing to do? Does anyone have
information as to how critical the HBOND terms are for GLYCAM?

Suggestions for either or both of the above connundra would be most welcome.

Thanks,

Tom.

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu