Hi all,
I just noticed that the copy of gaff.dat (the GAFF parameter
definition file) appearing in AmberTools 1.2 contains repeated
entries, but with these entries specifying different parameters. For
example, in the bond section:
cd-cd 418.3 1.429 SOURCE1 740 0.0058 0.0069
cd-cd 504.0 1.371 SOURCE3 523 same as cc-cd
A complete list of duplicate lines is attached below. This occurs
both in amber10/dat/leap/parm/gaff.dat and
amber10/dat/leap/gleap/gaff.dat.
Can anyone comment on which entry (the first or second) is correct,
and which entry LEaP would have been using (first or second)?
This issue appears to have been dealt with via a bugfix in the AMBER8 release:
http://ambermd.org/bugfixes/8.0/bugfix.28
but the buggy file (if this is indeed unintended) appears to have
crept back into the release versions somewhere along the line.
Applying this patch appears to fix the problems, but some patches are
rejected (also attached below).
This was also mentioned on the mail reflector as recently as June of this year:
http://archive.ambermd.org/200806/0289.html
but it appears to remain unrectified in the AmberTools 1.2
distribution put out in July.
Thanks go out to D. E. Shaw Research guys on this: I've been working
on AmberTools GAFF -> Desmond conversion scripts, and their forcefield
conversion tools detected and enumerated the duplicate entries in
gaff.dat.
Thanks!
John
--
John D. Chodera
QB3-Berkeley Fellow, University of California, Berkeley
url: http://www.choderalab.org
mobile: 415.867.7384
email: jchodera.berkeley.edu
--
Complete list of duplicated entries:
lines 336-337
cd-cd 418.3 1.429 SOURCE1 740 0.0058 0.0069
cd-cd 504.0 1.371 SOURCE3 523 same as cc-cd
lines 371-372
cf-cf 390.5 1.451 SOURCE1 66 0.0060 0.0060
cf-cf 562.4 1.338 SOURCE1 543 same as ce-cf
lines 392-393
ch-ch 494.2 1.377 SOURCE1 42 0.0000 0.0000
ch-ch 949.5 1.191 SOURCE1 80 same as cg-ch
lines 541-542
n1-n1 1221.7 1.124 SOURCE1 19 0.0000 0.0000
n1-n1 1365.7 1.100 SOURCE3
lines 653-654
nd-nd 486.8 1.379 SOURCE3 9 0.0145 0.0164
nd-nd 602.9 1.315 SOURCE3 9 same as nc-nd
lines 668-669
nf-nf 355.3 1.479 SOURCE3 19 same as ne-ne
nf-nf 738.6 1.257 SOURCE3 1 same as ne-nf
lines 789-790
pf-pf 240.7 2.092 SOURCE3 7 same as pe-pe
pf-pf 260.8 2.055 SOURCE3 1 same as pe-pf
lines 2076 and 4277
n1-n1-s 54.6 180.00 SOURCE3 1 0.0000 0.0000
n1-n1-s 54.6 179.96 HF/6-31G*
lines 1042 and 4285
n1-c1-n1 93.2 102.01 SOURCE0 1
n1-c1-n1 70.2 180.00 HF/6-31G*
Rejected patches from bugfix.28 for AMBER8 applied to gaff.dat in AmberTools-1.2
***************
*** 4224,4230 ****
c1-n1-o 62.6 179.95 HF/6-31G*
c1-n1-oh 62.9 174.31 HF/6-31G*
c1-n1-os 62.2 176.61 HF/6-31G*
! c1-n1-p1 83.5 172.83 HF/6-31G*
c1-n1-p3 68.7 173.51 HF/6-31G*
c1-n1-p4 68.7 173.64 HF/6-31G*
c1-n1-p5 71.5 177.28 HF/6-31G*
--- 4216,4222 ----
c1-n1-o 62.6 179.95 HF/6-31G*
c1-n1-oh 62.9 174.31 HF/6-31G*
c1-n1-os 62.2 176.61 HF/6-31G*
! c1-n1-p2 83.5 172.83 HF/6-31G*
c1-n1-p3 68.7 173.51 HF/6-31G*
c1-n1-p4 68.7 173.64 HF/6-31G*
c1-n1-p5 71.5 177.28 HF/6-31G*
***************
*** 5061,5075 ****
hc-o -c -oh 1.1 180. 2. Junmei et al.1999
hc-o -c -os 1.1 180. 2.
n2-c2-ca-n2 1.1 180. 2. dac guess, 9/94
- n2-ca-ca-c3 1.1 180. 2.
n2-ca-ca-n2 1.1 180. 2. dac guess, 9/94
- na-ca-ca-c3 1.1 180. 2.
na-n2-ca-n2 1.1 180. 2. dac, 10/94
X -X -X -X 1.1 180. 2. dac, 10/94
hw ow 0000. 0000. 4. flag
for fast water
- n na n2
MOD4 RE
h1 1.3870 0.0157 Veenstra et al JCC,8,(1992),963
--- 5046,5057 ----
hc-o -c -oh 1.1 180. 2. Junmei et al.1999
hc-o -c -os 1.1 180. 2.
n2-c2-ca-n2 1.1 180. 2. dac guess, 9/94
n2-ca-ca-n2 1.1 180. 2. dac guess, 9/94
na-n2-ca-n2 1.1 180. 2. dac, 10/94
X -X -X -X 1.1 180. 2. dac, 10/94
hw ow 0000. 0000. 4. flag
for fast water
MOD4 RE
h1 1.3870 0.0157 Veenstra et al JCC,8,(1992),963
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
to majordomo.scripps.edu
Received on Mon Oct 27 2008 - 05:08:40 PDT