AMBER Archive (2009)

Subject: Re: [AMBER] large resp charges for non-natural amino acid

From: Arturas Ziemys (
Date: Sat Jan 31 2009 - 11:03:09 CST

The send archives each alone, as AMBER banned due to the size. My
original message is below.

Arturas Ziemys wrote:
> Hi, Francois,
> Two *.tar.gz file are attched. Each contains minimization outputs of 3
> conformation as yous asked: HF/6-31G* Opt=Tight SCF(Conver=8). The NME
> group atom lines are adjusted.
> I believe that in GLA (alpha-carboxyl-glutamic acid) residues there is
> no choice to find a conformation without a salt bridge between -COO
> and -NH groups with reasonable low energy (comparing to the rest).
> Because there are two -COO groups and those are so close to backbone
> fragment - it is unavoidable. Charge -2 polarize all the local
> environment. I believe 3 conformation approach could be good
> representation in overall. NAP should have no issues.
> I don't think we will use Duan's FF, but should the conformations be
> optimized with implicit solvation ? Let me know if need anything
> regarding AA.
> With very best
> Arturas

> FyD wrote:
>> Arturas,
>>> I see. I would appreciate lot if you could run RED on the server as we
>>> still have no access it (my PI asked out legal office to approve the
>>> usage of RED ... - it takes time). I have two amino acids actually. I
>>> will attach two tar.gz files. If you could run both or one - great!
>> ok. We can run both amino acids (AA) in the same R.E.D. Server job.
>> Do you want also the N-terminal & C-terminal fragments, besides the
>> central one ? (This is the same amount of work: This will take just
>> ten minutes on R.E.D. Server). Do you want the version for the Duan
>> et al FF as well ?
>> May I ask you to ?
>> - change the atom order so that you have the following atom order in
>> your molecules: ACE-AA-NME
>> (to not have the NME residue in the middle of the AA residue)
>> - provide the QM geometry optimization output using
>> HF/6-31G* Opt=Tight SCF(Conver=8) or HF/6-31G** Opt=Tight
>> SCF(Conver=8)
>> (and not that single point job)
>> - finally think to the conformation(s) you want to use in the charge
>> derivation. As you saw the charge value of CA atom oscillate between
>> 0.1 - 0.3 in your case. Usually, one avoids H-bond when selecting a
>> conformation. You might consider using the ModRedundant mode in
>> Gaussian to block dihedrals/to remove those H-bonds...
>> regards, Francois
>> _______________________________________________
>> AMBER mailing list


Arturas Ziemys, PhD School of Health Information Sciences University of Texas Health Science Center at Houston 7000 Fannin, Suit 880 Houston, TX 77030 Phone: (713) 500-3975 Fax: (713) 500-3929

AMBER mailing list