Dear amber friends,
I recently noved to a new laptop (where I calculate energies after
cluster productive runs). In my former laptop, I had openSuSE 11.2 (64
bits), to which I compiled AmberTools 1.3 and Amber10, gcc version 4.4(.1 as
reported by configure)-4.2, all patches applied, following instructions at
http://archive.ambermd.org/201001/0150.html, and things ran as expected. To
my new laptop, I went to openSuSE 11.4 (64 bits), gcc is 4.5(.1 as reported
by configure)-19.1 (as reported by yast2). AmberTools compilation seems to
go fine, but at "test" I get the error:
"
make[1]: Entering directory `/usr/local/amber11/test/nab'
=====================================================
Running test to make dna duplex:
Can't open residue library /usr/local/amber11/dat/reslib/al_nucleic94.liib.
make[1]: *** [duplex_test] Error 1
make[1]: Leaving directory `/usr/local/amber11/test/nab'
make: *** [test.nab] Error 2
"
Exactly as pasted, "al_nucleic94.liib".
The closest I found to this was
http://archive.ambermd.org/201002/0267.html.
I tried a number of alternatives. One of them, the intel compiler, did
not compile (unfortunately this OS is said not to be supported for it,
although I followed the instructions at
http://software.intel.com/en-us/articles/installing-intel-compilers-on-opensuse/for
former versions, but without success (to compile amber, not compiler
installation).
I think the best I could do for while was to put openSuSE 11.2 in a
virtual machine, there it compiled nicely (as in the old laptop) and it
seems to work, although during a calculation I had a problem on " SA Bomb
in sa_arc(): Allocation aborted 0 0 0
0 5014" for a frame (memory limits in the virtual
machine?). But it does not seem to be a good idea to run amber under a
virtual machine. I made a (senseless?) test, simply copied the amber10 and
amber11 directories from the VM to the host OS, (hand) made the
(non-transferable) links and mk_crd_hg and sander seem to work (yes, they
could even calculate energias), but not tleap (so prm files have to come
from elsewhere). In fact, for this transferred tleap I get:
"
$AMBERHOME/bin/teLeap -I$AMBERHOME/dat/leap/prep -I$AMBERHOME/dat/leap/lib
-I$AMBERHOME/dat/leap/parm -I$AMBERHOME/dat/leap/cmd -s -f leaprc.ff99SB -f
leap.inp
-I: Adding /usr/local/amber11/dat/leap/prep to search path.
-I: Adding /usr/local/amber11/dat/leap/lib to search path.
-I: Adding /usr/local/amber11/dat/leap/parm to search path.
-I: Adding /usr/local/amber11/dat/leap/cmd to search path.
-s: Ignoring startup file: leaprc
-f: Source leaprc.ff99SB.
-f: Source leap.inp.
Welcome to LEaP!
Sourcing: /usr/local/amber11/dat/leap/cmd/leaprc.ff99SB
Log file: ./leap.log
Loading parameters: /usr/local/amber11/dat/leap/parm/parm99.dat
Reading title:
PARM99 for DNA,RNA,AA, organic molecules, TIP3P wat. Polariz.& LP
incl.02/04/99
Loading parameters: /usr/local/amber11/dat/leap/parm/frcmod.ff99SB
Reading force field modification type file (frcmod)
Reading title:
Modification/update of parm99.dat (Hornak & Simmerling)
Loading library: /usr/local/amber11/dat/leap/lib/all_nucleic94.lib
Segmentation fault
"
So again I understand the the problem is the all_nucleic94.lib file
(which is in there, in the due place, but...).
Some points: i) I would very prefer to use ambertools 1.3 and amber10
now yet the results are to be compared to several others runs with those
program versions; ii) I would prefer to keep the most updated version of
openSuSE.
Anyway, I think relating these problems might help development. Anyway
to overcome the error and compile natively under my (most uptodate)
distribution.
Suggestions?
Thanks,
Jorge
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 25 2011 - 07:00:03 PDT
