- Contemporary messages sorted: [ by date ] [ by thread ] [ by subject ] [ by author ] [ by messages with attachments ]

From: Jason Swails <jason.swails.gmail.com>

Date: Tue, 19 Apr 2011 10:36:12 -0700

Hello,

Your analysis is correct. The numbers to the right are the atom numbers.

The numbers to the left are the barrier height, phase shift, and

periodicity, respectively. Note that the barrier height is already divided

by the multiplicity factor (or however it's described in

http://ambermd.org/formats.html). Note that dihedral indexing/assignment

can get quite complicated depending on what terms are present in your force

field.

I have seen 2 prmtop files that, when diffed give a large number of

differences, but yield the same energies in sander. The bottom line is, I

think, that for different force fields, you can't expect the dihedral terms

to relate to one another in such a direct way.

As a demonstration of what I was talking about earlier, try running the

following script through tleap with ff99SB (or ff10):

sys = sequence {ACE ALA ALA ALA NME}

sys2 = copy sys

saveamberparm sys parm1 crd1

saveamberparm ssy2 parm2 crd2

If you diff parm1 and parm2, you'll see substantial differences, but they

will yield the same energy.

HTH,

Jason

On Tue, Apr 19, 2011 at 9:57 AM, david condon <dec986.gmail.com> wrote:

*> Hi All,
*

*>
*

*> I am checking dihedral parameters using rdparm for my LNA
*

*> tetranucleotide, and here are some results from rdparm's dihedral
*

*> command:
*

*>
*

*> for force field A:
*

*> 269: 0.000 0.00 2.0 :1.C2' :1.C1' :1.N1 :1.C2 (24,8,10,20)
*

*> 357: 0.000 0.00 2.0 :2.O4' :2.C1' :2.N9 :2.C4 (39,40,42,55)
*

*> 443: 0.000 0.00 2.0 :3.O4' :3.C1' :3.N9 :3.C4 (73,74,76,89)
*

*> 512: 0.000 0.00 2.0 :4.C2' :4.C1' :4.N1 :4.C2
*

*> (123,108,110,119)
*

*>
*

*> for force field B:
*

*> 251: 0.000 0.00 2.0 :1.C2 :1.N1 :1.C1' :1.O4' (20,10,8,7)
*

*> 367: 0.000 0.00 2.0 :2.O4' :2.C1' :2.N9 :2.C4 (39,40,42,55)
*

*> 458: 0.000 0.00 2.0 :3.O4' :3.C1' :3.N9 :3.C4 (73,74,76,89)
*

*> 514: 0.000 0.00 2.0 :4.C2 :4.N1 :4.C1' :4.O4'
*

*> (119,110,108,107)
*

*>
*

*> The two molecules are identical except for the force fields.
*

*>
*

*> Unfortunately the AMBER manual said nothing about how to read the
*

*> output. I thought that the four numbers at right were atom #s, but
*

*> the two molecules are identical. My question is, what are the 3
*

*> numbers on the left? I thought they were cosine terms but then the
*

*> different force fields are giving vastly different results, as they
*

*> were designed to.
*

*>
*

*> What are the 4 numbers on the right?
*

*>
*

*> thanks in advance for your time,
*

*> -Dave
*

*>
*

*> _______________________________________________
*

*> AMBER mailing list
*

*> AMBER.ambermd.org
*

*> http://lists.ambermd.org/mailman/listinfo/amber
*

*>
*

Date: Tue, 19 Apr 2011 10:36:12 -0700

Hello,

Your analysis is correct. The numbers to the right are the atom numbers.

The numbers to the left are the barrier height, phase shift, and

periodicity, respectively. Note that the barrier height is already divided

by the multiplicity factor (or however it's described in

http://ambermd.org/formats.html). Note that dihedral indexing/assignment

can get quite complicated depending on what terms are present in your force

field.

I have seen 2 prmtop files that, when diffed give a large number of

differences, but yield the same energies in sander. The bottom line is, I

think, that for different force fields, you can't expect the dihedral terms

to relate to one another in such a direct way.

As a demonstration of what I was talking about earlier, try running the

following script through tleap with ff99SB (or ff10):

sys = sequence {ACE ALA ALA ALA NME}

sys2 = copy sys

saveamberparm sys parm1 crd1

saveamberparm ssy2 parm2 crd2

If you diff parm1 and parm2, you'll see substantial differences, but they

will yield the same energy.

HTH,

Jason

On Tue, Apr 19, 2011 at 9:57 AM, david condon <dec986.gmail.com> wrote:

-- Jason M. Swails Quantum Theory Project, University of Florida Ph.D. Candidate 352-392-4032 _______________________________________________ AMBER mailing list AMBER.ambermd.org http://lists.ambermd.org/mailman/listinfo/amberReceived on Tue Apr 19 2011 - 11:00:03 PDT

Custom Search