[gmx-users] six member ring won't stay flat

[MD]

​Hi! You are completely right about the parametrization. I found the
pyrodoxal in charmm and I think I am going to start over to combine it with
lys which can reduce input from outside source to make charmm happy.
however I left the md.mdp run yesterday and this morning i found the methyl
group is fixed, the ring went back to flat... not quite sure what happened
between minimization and simulation.  ​but yes I wouldn't really trust the
results based on the discrepancies. Thanks for the advice :)

Ming

On Tue, Jan 16, 2018 at 7:37 AM, Justin Lemkul <jalemkul at vt.edu> wrote:

>
>
> On 1/16/18 6:38 AM, MD wrote:
>
>> Hi Justin,
>>
>> I got the itp and parameters of my side chain modified amino acid from
>> CHARMM-GUI and incorporated it into my protein structure, labeled with
>> HETATM. I made the atom types names consistent with charmm forcefield
>> which
>> I used with gromacs and made sure overall the parameters look decent for
>> now. After some fixing the grompp would run with no warnings, and I did a
>> quick energy minimization, but ended up with a distorted six member ring.
>> I
>> have the picture and my parameters attached. Your time is appreciated :)
>>
>> https://docs.google.com/document/d/1bjSq55HDLRsSVGqm5i0MRwI-
>> rgIesxy0QdIHm7tqTug/edit?usp=sharing
>>
>
> You have a number of problems. If that's simply pyridoxal phosphate linked
> with lysine, then you have lots of missing H atoms and the protonation
> state of your phosphate group is incorrect, at least with respect to normal
> physiological pH. You're getting a lot of distortion from a lot of places,
> not the least of which is that you should have a methyl group attached to
> the ring, not a methylene (=CH2), as that changes the conjugation entirely.
>
> Also, as I said before - *do not* mix CGenFF and standard CHARMM
> parameters. You can't just change around atom types until grompp warnings
> go away. What you're seeking to do is complicated and requires great care,
> otherwise you get garbage. There's no magic push-button here. You've got to
> do a thorough parametrization, including all the things I said before.
> Anything short of that, in this instance, is likely to fail. You can take
> initial guess charges from existing groups in CHARMM, without even going to
> CGenFF, as most of those groups should already be well described.
>
>
> -Justin
>
> --
> ==================================================
>
> Justin A. Lemkul, Ph.D.
> Assistant Professor
> Virginia Tech Department of Biochemistry
>
> 303 Engel Hall
> 340 West Campus Dr.
> Blacksburg, VA 24061
>
> jalemkul at vt.edu | (540) 231-3129
> http://www.biochem.vt.edu/people/faculty/JustinLemkul.html
>
> ==================================================
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/Support
> /Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>