On Mon, Jul 16, 2012, Yun Shi wrote:
>
> As a test, I used tleap to save .prmtop and .inpcrd files for RU3 (R-URACIL
> - with 5' - phosphate group and 3' - OH group) based on ff99SB.
>
> When I inspect the resulting .prmtop file with rdparm, I found the P
> -O5'-C5'-H5'1 and P -O5'-C5'-C4' dihedral parameters were
>
> 0.383 0.00 3.0
>
> But these two dihedrals have atom types as P -OS-CT-H1 and P -OS-CT-CT,
> which should match X -CT-OS-X (1.15 0.0 3.0) in parm99.dat.
They *do* match: the "X -CT-OS-X" line in parm99.dat is this:
X -CT-OS-X 3 1.15 0.0 3.
You have to divide the 1.15 by 3 (the "idiv" value) to get the 0.383 value
that rdparm reports.
> this 0.383 force constant is from one term of a combination like
> CT-CT-OS-CT/C/N* (0.383 0.0 -3.0) in parm99.dat.
This is also consistent: the idiv value for CT-CT-OS-CT is 1, so pk/idiv is
the same as for the generic (wild-card) torsion.
>
> I also found similar discrepancy for O1P-P -O5'-C5' (atom type: O2-P
> -OS-CT), which was recognized as only the first term of OS/OH-P -OS-CT
> (0.25 0.0 -3.0) in the prmtop file. But I assume it should use X -OS-P -X
> (0.75 0.0 3.0).
Same explanation as above: don't ignore the idiv values.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jul 16 2012 - 12:00:03 PDT
