From: Michael LeVine (mlevine_at_wesleyan.edu)
Date: Mon Jul 26 2010 - 10:10:28 CDT
Hi all,
I'm currently trying to run some FEP calculations for a L->G mutant. I
continue to get the following error:
CAN'T FIND IMPROPER PARAMETERS FOR CT1 H C CT1
I used alchemify to make sure I have all my exclusion data for the FEP run,
but that doesn't change a thing. Here is my topology for the hybrid residue:
RESI G2L 0.00
GROUP
ATOM N NH1 -0.47 ! |
ATOM HN H 0.31 ! N-H
ATOM CA CT2 -0.02 ! |
ATOM HA1 HB 0.09 ! |
ATOM HA2 HB 0.09 ! HA1-CA-HA2
GROUP ! |
ATOM C C 0.51 ! |
ATOM O O -0.51 ! C=O
! |
GROUP
ATOM CAL CT1 0.07 ! | HB1 CD1--HD13
ATOM HAL HB 0.09 ! | | /
GROUP ! HA-CA--CB--CG-HG
ATOM CBL CT2 -0.18 ! | | \
ATOM HB1L HA 0.09 ! | HB2 CD2--HD23
ATOM HB2L HA 0.09 ! O=C | \
GROUP ! | HD21 HD22
ATOM CGL CT1 -0.09
ATOM HGL HA 0.09
GROUP
ATOM CD1L CT3 -0.27
ATOM HD11L HA 0.09
ATOM HD12L HA 0.09
ATOM HD13L HA 0.09
GROUP
ATOM CD2L CT3 -0.27
ATOM HD21L HA 0.09
ATOM HD22L HA 0.09
ATOM HD23L HA 0.09
BOND N HN N CA C CA
BOND C +N CA HA1 CA HA2
BOND CBL CAL CGL CBL CD1L CGL CD2L CGL
BOND N CAL C CAL
BOND CAL HAL CBL HB1L CBL HB2L CGL HGL CD1L HD11L
BOND CD1L HD12L CD1L HD13L CD2L HD21L CD2L HD22L CD2L HD23L
DOUBLE O C
IMPR CAL HN C CAL
IMPR N -C CA HN C CA +N O
CMAP -C N CA C N CA C +N
CMAP -C N CAL C N CAL C +N
DONOR HN N
ACCEPTOR O C
PATCHING FIRS GLYP
I've tried this with and without the patching line and nothing changed. Is
there an improper I need to add? I have all the improper data from the
original topologies for both. I can't find much on this issue except for the
MUTATOR plug-in page warning against doing it unless you know what you are
doing. Too late for that, I guess.
Also, I have P->S mutant which has had a strange issue. I'm running a
PROTEIN:LIGAND FEP and PROTEIN FEP so I can calculate the dG for the
mutation on ligand binding. However, while the PRO:LIG FEP runs fine with
some initial minimization, the unbound run starts with infinite VDW energy.
Upon inspection of the PDB, both the alpha-carbons and nitrogens have the
same coordinates; they are not the same atom type however, so I need them to
be unique atoms for the run. This occurs in my bound structure as well, but
the energy does not begin with infinite VDW and minimization for 5000 steps
fixes the high initial energy fine. How can I overcome this issue, and why
would the VDW be infinite for the free and finite for the bound? The atom
coordinates are varied at this part of the protein by over 1 A between the
bound and free proteins. I was considering forcing the atoms away from each
other and hoping that despite the initial strain, when the system minimized
they would have unique coordinates and I wouldn't have the VDW issue. Is
this a plausible plan, or are there long term implications for the
calculations? Thanks for the help.
Mike
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:55:59 CST