Re: AMBER: respin file

From: M. L. Dodson <mldodson.houston.rr.com>
Date: Sun, 26 Nov 2006 12:05:33 -0600

On Sunday 26 November 2006 11:12, Fenghui Fan wrote:
> Than you very much for your reply.
>
> As for the CYM.pdb structure mentioned in your
> protocol, it should be not be neutral, because it
> lacks the H in the SH and one H in the N-terminal NH2.
>

This is probably why the antechamber tools cannot figure out the
bond types (as you indicated in an earlier posting). This will
not be a closed shell QM calculation. I would even go so far to
say this approach is never going to work for that reason.
Proteins, even Zn-containing proteins, are very much more likely
than not to be closed shell species. To generate new residues for
amber you should do closed shell calculations, then handle partial
charges on the atoms to be removed to form the final residue
definition (at the sites of covalent bonds to be created in LEaP)
by constraining their charge to zero in the resp fit. Maybe REDII
can handle this (mostly) automatically?

> Somebody treat the bond between the SG of CYM and the
> ZN as a covalent bond. I have tried this way, the
> situation is that the prepin file of the CYM-ZN.pdb
> lackes the information of the ZN and the last two
> atoms (I think the last 2 atoms should be ZN), thus in
> the frcmod file, there is no information on the ZN
> related bonds.

There is a lot more to handling Zn in amber than just getting all
the programs to "work". Please review the mailing list archives
for postings (going all the way back to at least 2001) for
discussions about how to handle Zn ions in amber calculations.
You will also have to read original literature and develop a
thorough understanding of the underlying chemistry. You might
search for papers from Dr. K. Merz's lab; also:

Y-P Pang, Proteins 45:183-189 (2001)
Biochemistry. 2003 Aug 5;42(30):8945-56)
Protein Eng. 2003 May;16(5):341-50
Prot. Science, v9, pp. 1857
and an article by Stote and Karplus in Proteins a few years back

These references are from a simple search for "Zn parameters" in
the amber mailing list archives.

"Correct" treatment of Zn in the amber force field is definitely
not good choice for a beginner wanting to learn about using the
amber suite of tools. Read a selection of the (many) papers from
the Merz lab to get a feeling of the complexities that can arise.
Of course, these papers are treating a catalytic Zn. The Zn in,
e.g., a Zn finger protein, will probably allow a less detailed
approximation of Zn chemistry, but without a good understanding of
the chemistry issues it will be difficult to choose the correct
level of approximation for your planned studies. I suggest
spending the time to understand these issues will move you forward
in your studies faster than just plunging ahead trying to get the
programs to "work". Especially if your definition of "working" is
just getting the programs to execute without printing an error
message.

Bud Dodson

>
> For your protocol, if we treat ZN separately, that
> means we will have a PDB contains the coordinates of
> only ZN, then we can get the ZN library file, but how
> about its prepin file and the its frcmod file?
>
> The other question is about the command you mentioned:
>
>
> antechamber -i CYM.pdb -fi pdb -o cym.in -fo gcrt
>
> What will be the charge type of this command?
>
> I am looking forward to getting your reply.
>
> Best regards.
>
> Fenghui Fan
>
> --- Ilyas Yildirim <yildirim.pas.rochester.edu> wrote:
> > The .pdb file u have attached have 2 unique
> > residues, one is called CYM
> > and the other one is Zn. What u have to do is erase
> > the Zn residue, and
> > save the file (for example, as CYM.pdb). Then,
> >
> > antechamber -i CYM.pdb -fi pdb -o cym.in -fo gcrt
> >
> > This will create a file 'cym.in', which is a
> > gaussian input file. It will
> > do both the gaussian optimization + MEP calculation
> > (aka ESP data point
> > calculation). After finishing this gaussian
> > calculation, u can -again- use
> > antechamber to create (for instance) a .mol2 file,
> > which will let u load
> > this structure into LeAP.
> >
> > For the Zn, u need to use other parameters (if there
> > are any parameter
> > set defined for Zn). In this protocol, I assumed
> > that the structure
> > CYM.pdb is neutral. If it is not, u have to modify
> > cym.in file (For more
> > information on how to do it, check out the gaussian
> > manual, which can be
> > found at http://www.gaussian.com/g_ur/g03mantop.htm
> > - this is a very
> > useful online manual if u use gaussian often).
> >
> > You might use the program R.E.D. II for this resp
> > charge calculation. For
> > more information on R.E.D., check out the following
> > website:
> >
> > http://www.u-picardie.fr/labo/lbpd/RED/
> >
> > Best,
> >
> > PS: I might be missing something here. I assumed
> > that u are trying to find
> > the resp charges. The above explanation is to find
> > the resp charges. After
> > getting these charges for CYM.pdb atoms, u can use
> > antechamber and create
> > a .mol2 file, which will have the atom type info in
> > it.
> >
> > On Sat, 25 Nov 2006, Fenghui Fan wrote:
> > > You are right.
> > >
> > > The thing is that during my ac file preparation,
> >
> > the
> >
> > > antechamber cannot get the correct bond type, thus
> >
> > I
> >
> > > use -j 0. However when I do as you suggets, the
> >
> > prepin
> >
> > > file and the frcmod file are almost empty, that
> >
> > means,
> >
> > > there is nothing related to each atom or each
> >
> > bond.
> >
> > > Followng is my original PDB file:
> > >
> > > -------
> > > HEADER Structure from MOLMOL
> > > COMPND CYM_ZN001
> > > ATOM 1 N CYM 33 -2.632 -0.454
> >
> > 0.499
> >
> > > 1.00 0.00
> > > ATOM 2 H CYM 33 -2.424 -0.481
> >
> > 1.487
> >
> > > 1.00 0.00
> > > ATOM 3 CA CYM 33 -1.572 0.022
> >
> > -0.433
> >
> > > 1.00 0.00
> > > ATOM 4 HA CYM 33 -1.971 0.801
> >
> > -1.066
> >
> > > 1.00 0.00
> > > ATOM 5 CB CYM 33 -0.391 0.570
> >
> > 0.372
> >
> > > 1.00 0.00
> > > ATOM 6 3HB CYM 33 -0.110 -0.143
> >
> > 1.130
> >
> > > 1.00 0.00
> > > ATOM 7 2HB CYM 33 -0.677 1.500
> >
> > 0.840
> >
> > > 1.00 0.00
> > > ATOM 8 SG CYM 33 1.014 0.862
> >
> > -0.733
> >
> > > 1.00 0.00
> > > ATOM 9 C CYM 33 -1.104 -1.150
> >
> > -1.298
> >
> > > 1.00 0.00
> > > ATOM 10 O CYM 33 -0.671 -2.169
> >
> > -0.798
> >
> > > 1.00 0.00
> > > TER
> > > ATOM 1 Zn Zn 55 2.800 1.527
> >
> > 0.560
> >
> > > 1.00 0.00
> > > TER
> > > END
> > >
> > > ----------
> > >
> > > Will you please tell me the command how can I get
> >
> > the
> >
> > > right gaussian input file (just call it input.ac)?
> >
> > My
> >
> > > way is use -j 0, but it cannot give the correct
> >
> > prepin
> >
> > > and frcmod. But if I delete -j 0, the input.ac
> >
> > cannot
> >
> > > be got by antechamber as the unrecognizable
> >
> > bondtype.
> >
> > > I am looking forward to getting your reply.
> > >
> > > Best regards.
> > >
> > > Fenghui Fan
> > >
> > > --- Ilyas Yildirim <yildirim.pas.rochester.edu>
> >
> > wrote:
> > > > I guess you are confusing yourself with the
> > > > definitions. Let me explain a
> > > > little bit more on what I tried to say in my
> > > > previous email.
> > > >
> > > > 1. You have a structure. Let us call it as
> >
> > x.pdb.
> >
> > > > 2. You use an ab initio program (like gaussian)
> >
> > to
> >
> > > > optimize the structure.
> > > >
> > > > 3. You use an ab initio program (again) to get
> >
> > the
> >
> > > > ESP data points (PS:
> > > > The 2nd and 3rd steps can be combined. If you
> >
> > have
> >
> > > > created the gaussian
> > > > input file for your structure x.pdb using
> > > > antechamber, this is the case).
> > > >
> > > > 4. After finishing the second phase of the
> >
> > gaussian
> >
> > > > calculation (which is
> > > > getting the ESP Data points), u use antechamber
> >
> > to
> >
> > > > get the resp charges.
> > > > When u use antechamber, it will do all the steps
> > > > automatically.
> > > >
> > > > As I said before, I assume that u are trying to
> >
> > get
> >
> > > > the resp charges for
> > > > just 1 structure (meaning u dont need to deal
> >
> > with
> >
> > > > equivalencing). Hope
> > > > this helps.
> > > >
> > > > Best,
> > > >
> > > > On Sat, 25 Nov 2006, Fenghui Fan wrote:
> > > > > Thank you very much.
> > > > >
> > > > > First I get the gaussian output file, then I
> >
> > use
> >
> > > > the
> > > >
> > > > > gaussian output file to get the esp file, the
> >
> > I
> >
> > > > start
> > > >
> > > > > to get the resp charge, but I cannot get the
> >
> > qout.
> >
> > > > > I will refer to your e-mail and try to make it
> > > > > workable. If you have any comments, please
> >
> > tell
> >
> > > > me.
> > > >
> > > > > Best regards.
> > > > >
> > > > > Fenghui Fan
> > > > >
> > > > >
> > > > >
> > > > > --- Ilyas Yildirim
> >
> > <yildirim.pas.rochester.edu>
> >
> > > > wrote:
> > > > > > Did you try to use antechamber with the
> >
> > gaussian
> >
> > > > > > output file as the input
> > > > > > file to get the resp charges? I think, if
> >
> > you
> >
> > > > have
>
> === message truncated ===
>
>
>
>
> _________________________________________________________________________
>___________ Do you Yahoo!?
> Everyone is raving about the all-new Yahoo! Mail beta.
> http://new.mail.yahoo.com
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu

-- 
M. L. Dodson
Email:	mldodson-at-houston-dot-rr-dot-com
Phone:	eight_three_two-56_three-386_one
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Nov 29 2006 - 06:07:09 PST
Custom Search