AMBER Archive (2009)

Subject: [AMBER] A bug in LEaP II ?

From: FyD (fyd_at_q4md-forcefieldtools.org)
Date: Sun Oct 25 2009 - 07:58:36 CDT

Dr Case, dear All,

Around two weeks ago, I sent the email below, about a possible bug in
t/x-LEaP:
It was about using the "P1.off" FF library. No answer.

I did more tests & it looks like when using the anomer beta of P1.off
(i.e. Q1.off) there is no bug. Please, find attached (again) this
P1.off FF library (anomer alpha) and the new Q1.off (beta anomer).
When using the beta-anomer Q1.off in the LEaP script below, all work
just perfectly independently of using the "copy" command; (obviously
after replacing "P" by "Q" in this script). However, we still have the
problem reported when using P1.off: the glycosidic connection is
inverted and the first unit cycle is modified.

The "diff Q1.off P1.off" command provides what one would expect: name
differences between P1 & Q1 + expected Cartesian coordinate
differences (for the 2 atoms inverted because of using the "flip
chirality" command).

This is obviously not a general problem related to alpha anomers since
we have other FF libraries describing alpha anomers and we do not have
this problem...

Any idea of the problem ?
Do you think a correction will be available in the AmberTools 1.3 ?

regards, Francois

----- Forwarded message from fyd_at_q4md-forcefieldtools.org -----
     Date: Sun, 11 Oct 2009 14:08:37 +0200
     From: FyD <fyd_at_q4md-forcefieldtools.org>
Reply-To: AMBER Mailing List <amber_at_ambermd.org>
  Subject: [AMBER] A bug in LEaP ?
       To: AMBER Mailing List <amber_at_ambermd.org>

Dear All,

We use t/x-LEaP to build FF libraries (last version available - under
Fedora 10 64 bits; the same problem appends on Ubuntu). If we start
from the attached "P1.off" FF lib. & we run the LEaP script provided
below using:

xleap -f leap.cmd

some di-units T1, T2, T3 & T4 are wrong. In fact, odd T- FF libs are
wrong & even T- FF libs are correct. This means that starting from the
"P1.off" FF lib the "copy" command can solve or break a di-unit.
Thus, it looks like LEaP does not generate a constant FF lib for this
"P1.off" file and it is related to the order of the atom connectivities.

I hope I miss something here, because potentially prmtop files are wrong...

Any idea for this problem ?

Thanks, regards, Francois

------------------------

verbosity 2
loadoff P1.off

P2 = copy P1
saveoff P2 P2.off
P3 = copy P2
saveoff P3 P3.off
P4 = copy P3
saveoff P4 P4.off

T1 = sequence { P1 P1 }
T2 = sequence { P2 P2 }
T3 = sequence { P3 P3 }
T4 = sequence { P4 P4 }

savemol2 T1 T1.mol2 1
savemol2 T2 T2.mol2 1
savemol2 T3 T3.mol2 1
savemol2 T4 T4.mol2 1

----- End forwarded message -----


  • text/plain attachment: P1.off

  • text/plain attachment: Q1.off

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber