[gmx-users] LINCS warning in md run

[Justin Lemkul]

On 9/13/12 4:22 AM, reisingere at rostlab.informatik.tu-muenchen.de wrote:
> Hi Justin,
>
> this atom is the CG atom of the phosphate I added to the protein.
> But how can I find out by what it is punished?
>

Visualization.

> This is the topology for this residue in the aminoacid.rtp file. I added
> it manually. The parameter are from the amber parameter database
> (http://www.pharmacy.manchester.ac.uk/bryce/amber/):
>
> [ TYP ]
>   [ atoms ]
>       N    N           -0.516300    1
>      CA    CT           0.275503    2
>      HA    H1           0.008223    3
>      CB    CT          -0.354052    4
>     HB1    HC           0.110326    5
>     HB2    HC           0.110326    6
>      CG    CA           0.119728    7
>     CD1    CA          -0.198938    8
>     HD1    HA           0.137143    9
>     CE1    CA          -0.284884   10
>     HE1    HA           0.177179   11
>      CZ    C            0.452616   12
>      OH    OS          -0.534452   13
>       H    H            0.293600   14
>     CE2    CA          -0.284884   15
>     HE2    HA           0.177179   16
>     CD2    CA          -0.198938   17
>     HD2    HA           0.137143   18
>       C    C            0.536600   19
>       O    O           -0.581900   20
>       P    P            1.393213   21
>     OP1    OH          -0.752821   22
>     OP2    O2          -0.822464   23
>     OP3    O2          -0.822464   24
>     H1P    HO           0.423316   25
>   [ bonds ]
>       N     H
>       N    CA
>      CA    HA
>      CA    CB
>      CA     C
>      CB   HB1
>      CB   HB2
>      CB    CG
>      CG   CD1
>      CG   CD2
>     CD1   HD1
>     CD1   CE1
>     CE1   HE1
>     CE1    CZ
>      CZ    OS
>      CZ   CE2
>      OS     P
>     CE2   HE2
>     CE2   CD2
>     CD2   HD2
>       C     O
>      -C     N
>       P   OP1
>       P   OP2
>       P   OP3
>     OP1   H1P
>
> [ impropers ]
>      -C    CA     N     H
>      CA    +N     C     O
>      CG   CE2   CD2   HD2
>      CZ   CD2   CE2   HE2
>     CD1    CZ   CE1   HE1
>      CG   CE1   CD1   HD1
>     CD1   CD2    CG    CB
>     CE1   CE2    CZ    OH
>
>
> Additionally I had to add an entry in the ffbonded file in the part [
> dihedraltypes ]
>
> CA  CA  C   OS       4      180.00     4.60240     2    ;
>
> Here I am not completely sure about the entry because I could not find a
> source for this combination of atoms.
> But there is already an entry for the combination
> CA  CA  C OH
> Which has the same parameters.
> Could this be the problem?
>
> And if it could be, can you please give me hint where I can find
> parameters for it? Because I really could not find it.
>

Potentially - if you haphazardly assign parameters, you may get unexpected 
results.  There is a cited reference for the TYP parameters at the website 
above.  I would suggest looking there first, and if there is nothing listed, 
then contact the corresponding author of the study that first used/derived the 
parameters.

Additionally, if you freeze a part of the protein's structure, you prevent any 
existing clashes from being resolved.  That may be the root of the issue.

-Justin

-- 
========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================