Dear Amber community members,
I wrote a week ago about some trouble with organizing my files for parametrization of a laccase metal cluster using MCPB. Luckily, I've managed to iron those out, and now I am at the stage of the "small model" as described in the MCPB tutorial. Sadly, the gaussian optimization is doing pretty crazy things.
So, first of all, the laccase metal cluster is a tetranuclear copper center, where all coppers (at least in the state I am working with) should be CuII. Three of the coppers form a trinuclear center, while the fourth one is some 13A away, and is by itself. The trinuclear center is complexed by six histidines and two hydroxide ions, while the mononuclear one is complexed by two histidines and one cysteine.
Up until now, I have tried two strategies:
a) optimize both clusters together (after all, they should interact with each other during protein activity).
b) Optimize each cluster by itself.
In case a) convergence is very slow (it actually has not happened yet, after one week), but the worrying bit is that the two clusters separate in space. I don't think that is what they should do, is it?
In case b) the mononuclear cluster optimizes within a couple of hours, and looks very good, but the trinuclear one fails after ~2-3 days, with the following error message:
Berny optimization.
Using GEDIIS/GDIIS optimizer.
Bend failed for angle 64 - 123 - 120
Tors failed for dihedral 58 - 64 - 123 - 120
Tors failed for dihedral 67 - 120 - 123 - 64
Tors failed for dihedral 121 - 120 - 123 - 64
Tors failed for dihedral 122 - 120 - 123 - 64
FormBX had a problem.
Error termination via Lnk1e in /pkg/chem/gaussian/g09d01/g09/l103.exe at Wed Jul 3 06:27:09 2013.
Job cpu time: 31 days 16 hours 57 minutes 44.0 seconds.
File lengths (MBytes): RWF= 1314 Int= 0 D2E= 0 Chk= 108 Scr= 1
Error: segmentation violation
rax 0000000000000000, rbx ffffffffffffffff, rcx ffffffffffffffff
rdx 0000000000006a2e, rsp 00007fff96d640c8, rbp 00007fff96d646b0
rsi 000000000000000b, rdi 0000000000006a2e, r8 0000000000000000
r9 00002ab5d28a6e10, r10 ffffffffffffffff, r11 0000000000000202
r12 0000000000000000, r13 0000000000000000, r14 00007fff96d646f8
r15 00000000000003e6
--- traceback not available
Which, I suppose, means that some of the bonds are in a very awkward position.
I am really quite a newbee on this whole metalo-cluster modeling thing, so it would be absolutely great if someone could give me a hand!
Thank you very much in advance!
Manuel Maestre Reyna
Academia Sinica
Institute of Biological Chemistry
128, Academia Road Sec. 2, Nankang, Taipei 115, Taiwan
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jul 03 2013 - 22:00:02 PDT
