Dear AMBER users,
I'm not sure if this is the right place to post this question, but I would
appreciate any suggestions from you.
I'm going to use MOPAC509mn in conjunction with antechamber on a
linux(RedHat 8.0) machine (IBM note PC, PentiumIV).
I have successfully compiled the source codes using gcc/g77, although a
slight modifications were needed on timing routines.
However, all the test jobs provided with the software give incorrect heat
of formation.
For example, while test1 should give heat of formation of -36.87 kcal, but
the value I get is -20137.13673 kcal.
In addition, the program shows aberrant behaviors(very few optimization
steps compared with correct ones, etc.) before reaching a convergence.
Strangely, an executable on SGI compiled from virtually the same source
codes(except timing routines), using SGI 's
authentic f77 compiler, gives correct answers.
I tried a number of compiling options, including the one suggesting by
James Caldwell(shown below), but none of them works.
Any of you had a similar experience?
Thanks in advance,
Takanori Kanazawa
----------------------------------------------------------------------------------------------
From: James W. Caldwell (caldwell_at_heimdal.compchem.ucsf.edu)
Date: Tue Sep 24 2002 - 10:00:20 CDT
I just noted that the recommended mopac for Amber7/antechamber use
only has a make file for the (not free) Portland group fortran compiler.
I initially tried compiling it with the (free) Intel Fortran compiler
on my AMD systems and found that altho it does compile NONE of the
test cases pass. Even with no code optimization. NOTE: ifc does
work very well on Amber.
I tried G77 and all is fine (I couldn't get a make file to work with
g77 myself). Here is the make file in the style of the others which come
with the code, supplied by Dave Case, that will compile Mopac509mn
using the standard Linux G77 compiler:
# This is a sample makefile which compiles and links
# MOPAC 509mn on LINUX
# Set the compiler and loader commands.
#
FC = g77
FFLAGS= -O2
LD= g77
LDFLAGS=
MOPAC_MAIN = mopac_main.o
MOPAC = m509_mod.o m509_unmod.o betsrp.o chgmp2.o
# MOPAC machine-dependant routines
MMD = m509_mdep_ws.o m509_ef_ws.o
# General machine-dependant routines
GMD = second_linux.o date_linux.o fromblas.o
CFI = second1.o dateclock.o
mopac509: $(CFI) $(MOPAC) $(MOPAC_MAIN) $(MMD) $(GMD)
$(LD) -o mopac509mn $(MOPAC) $(MOPAC_MAIN) $(MMD) $(GMD) $(CFI)
$(LDFLAGS)
$(MOPAC) $(MOPAC_MAIN) $(MMD) : SIZES.i
clean :; /bin/rm *.o
--
----------------------------------------------------------------------------
James W. Caldwell (voice) 415-476-8603
Department of Pharmaceutical Chemistry (fax) 415-502-1411
Mail Stop 0446 (email) caldwell_at_heimdal.ucsf.edu
513 Parnassus Avenue
University of California
San Francisco, CA 94143-0446
----------------------------------------------------------------------------
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Fri Jul 25 2003 - 01:53:00 PDT
