AMBER Archive (2009)

Subject: Re: [AMBER] Problem about adding a small residue to DNA

From: FyD (fyd_at_q4md-forcefieldtools.org)
Date: Mon Sep 07 2009 - 04:53:14 CDT


Dear Yikan,

I have followed your questions & the corresponding answers in the past
few days. I would like to propose you an alternative way. I think/hope
this will help you.

First, GAFF (atom types with lower case) is generally used for organic
structures/ligands of structures, while Amber99SB (with the additional
FF parameters: frcmod.parmbsc0 & frcmod.ff99SB) for nucleic acids &
proteins (atom types with capital letters). Second, Antechamber has
been designed for organic molecules & works (I mean without errors)
for whole molecules (not molecular fragments).

This means in your case, you have a nucleic acid. You should use
Amber99SB in particular to take into account the last FF parameters
available (i.e. these described in frcmod.parmbsc0). However, we could
imagine using GAFF for a non-bounded ligand to your nucleic acid. In
your case, because TMS is physically connected to your nucleic acid, I
suggest you to use capital FF atom types related to the Amber99SB FF
for TMS and not GAFF ones.

What you did:
- When you load GAFF & save the prmtop/prmcrd for your TMS fragment,
it obviously works since TMS is isolated (not yet connected to your
oligonucleotide). However, when you save the prmtop/prmcrd of your
whole oligonucleotide containing your TMS fragment, it does not work
since you need new "hybrid" FF parameters (with both lower cases &
capitals):

Examples:
Building bond parameters.
Could not find bond parameter for: o - P
Could not find bond parameter for: o - P
Building angle parameters.
Could not find angle parameter: o - P - O2
Could not find angle parameter: o - P - OS
etc...
Building proper torsion parameters.
   ** No torsion terms for h1-c3-o-P
   ** No torsion terms for c3-o-P-O2
etc...

- If you do not pay attention to the additional FF parameters (you
have to provide now) you will not use these defined in
frcmod.parmbsc0, and more generally you will not use the FF parameters
specific/required to/for nucleic acids.

- Finally, you used Antechamber to generate charges & GAFF atom types
for a molecular fragment. As I already said, Antechamber was designed
"for organic molecules & for whole molecules". Consequently, when you
use Antechamber for a molecular fragment you get hidden errors (which
are not obvious for a new user). I think some FF atom types given by
Antechamber for your TMS fragment are wrong and in particular that
linking oxygen which should have the "os" atom type (or better OS atom
type).

Is there other errors ? I did not check carefully: Thus, what is below
is wrong (- even if you do decide to use GAFF):
Could not find bond parameter for: o - P
Could not find bond parameter for: o - P
Building angle parameters.
Could not find angle parameter: o - P - O2
Could not find angle parameter: o - P - OS
etc..
Building proper torsion parameters.
   ** No torsion terms for h1-c3-o-P

It should be:
Could not find bond parameter for: os - P
Could not find bond parameter for: os - P
Building angle parameters.
Could not find angle parameter: os - P - O2
Could not find angle parameter: os - P - OS
etc...
Building proper torsion parameters.
   ** No torsion terms for h1-c3-os-P

or better:
Could not find bond parameter for: OS - P
Could not find bond parameter for: OS - P
Building angle parameters.
Could not find angle parameter: OS - P - O2
Could not find angle parameter: OS - P - OS
etc...
Building proper torsion parameters.
   ** No torsion terms for H1-CT-OS-P

Consequently, I propose you:
- To use R.E.D. to get RESP charges for your TMS fragment (using
multiple conformation & multiple orientation approach). R.E.D.
provides you charges embedded in a FF library in the Tripos mol2 file
format. FF atom types are not provided to let users decide which ones
to add (and which FF to select). See tutorials @
http://q4md-forcefieldtools.org/Tutorial/.

- You learn what should be the FF atom types (capital letters) in your
TMS molecule using the article Cornell et al, JACS 1995, 117, 5179 -
Table 1 page 5182. & you read the FF atom type description available @
$AMBERHOME/dat/leap/parm/parm99.dat.
[In your TMS case, this is quite simple: you only need the CA, HA atom
types for the aromatic part; C, O, N, H for the peptidic bond and OS
for ether & ester linkages. Metyelenes bears CT & HC (or H1 if there
is one withdrawing atom connecting the carbon)]

- You manually add these FF atom types with capital letters in the FF
library generated by R.E.D. (you replace the chemical symbol by the FF
atom types: in 6th column in section "@<TRIPOS>ATOM").

- You load this FF library in LEaP after having loaded the AmberFF99SB
FF related files, and you save the prmtop/prmcrd files. LEaP will
complain about the missing FF parameters (if any). In your case, the
number of missing FF parameters should not be that important: I would
even bet that no missing FF parameters will be found for this TMS
fragment (it depends on the choice you are going to do for the central
double bond). Then, you create//provide the missing FF parameters (if
required in your case) in LEaP in a frcmod file manually built (with
capital letters) in a second LEaP run, and you will be able to save
the prmtop/prmcrd files for your entire oligonucleotide.

In these conditions, you use a single 'type' of force field atom types
(capital letters), a single force field (Amber99SB) specific to your
oligonucleotide and you have RESP charges derived as in the Amber
force field topology database (Cieplak et al. J. Comput. Chem. 1995,
16, 1357).

This is the way I would follow: And for sure, it will works... I hope
this helps. Sorry for this long email.

regards, Francois

Quoting Yikan Chen <yc2r_at_virginia.edu>:

> OK,I still do not know why I copy a wrong system message there.... When I
> ran saveamberparm instruction, the message shown like this (other
> instructions same as my former mail):
>
>> test=loadpdb a.pdb
> Loading PDB file: ./a.pdb
> Illegal CONECT record in pdb file
> total atoms in file: 300
> Leap added 182 missing atoms according to residue templates:
> 180 H / lone pairs
> 2 unknown element
>> saveamberparm test newff.prmtop newff.inpcrd
> Checking Unit.
> WARNING: The unperturbed charge of the unit: -14.000000 is not zero.
> -- ignoring the warning.
> Building topology.
> Building atom parameters.
> Building bond parameters.
> Could not find bond parameter for: o - P
> Could not find bond parameter for: o - P
> Building angle parameters.
> Could not find angle parameter: o - P - O2
> Could not find angle parameter: o - P - O2
> Could not find angle parameter: o - P - OS
> Could not find angle parameter: c3 - o - P
> Could not find angle parameter: o - P - O2
> Could not find angle parameter: o - P - O2
> Could not find angle parameter: o - P - OS
> Could not find angle parameter: c3 - o - P
> Building proper torsion parameters.
> ** No torsion terms for h1-c3-o-P
> ** No torsion terms for h1-c3-o-P
> ** No torsion terms for c3-o-P-O2
> ** No torsion terms for c3-o-P-O2
> ** No torsion terms for c3-o-P-OS
> ** No torsion terms for c3-c3-o-P
> ** No torsion terms for h1-c3-o-P
> ** No torsion terms for h1-c3-o-P
> ** No torsion terms for c3-o-P-O2
> ** No torsion terms for c3-o-P-O2
> ** No torsion terms for c3-o-P-OS
> ** No torsion terms for c3-c3-o-P
> Building improper torsion parameters.
> old PREP-specified impropers:
> <TMS 1>: C3 C2 C1 HE2
> <TMS 1>: C4 C1 C2 HE3
> <TMS 1>: C8 C7 C4 C2
> <TMS 1>: C4 C11 C7 HB6
> <TMS 1>: C14 C7 C11 O20
> <TMS 1>: C12 C11 C14 O21
> <TMS 1>: C14 C8 C12 O22
> <TMS 1>: C12 C4 C8 HB7
> <TMS 1>: C6 C5 C3 C1
> <TMS 1>: C3 C10 C6 HC5
> <TMS 1>: C6 C13 C10 HD9
> <TMS 1>: C3 C9 C5 HC4
> <TMS 1>: C5 C13 C9 HD8
> <TMS 1>: C15 C10 C13 C9
> <TMS 1>: C13 N1 C15 O1
> <TMS 1>: C16 C15 N1 H1
> <TMS 8>: C3 C2 C1 HE2
> <TMS 8>: C4 C1 C2 HE3
> <TMS 8>: C8 C7 C4 C2
> <TMS 8>: C4 C11 C7 HB6
> <TMS 8>: C14 C7 C11 O20
> <TMS 8>: C12 C11 C14 O21
> <TMS 8>: C14 C8 C12 O22
> <TMS 8>: C12 C4 C8 HB7
> <TMS 8>: C6 C5 C3 C1
> <TMS 8>: C3 C10 C6 HC5
> <TMS 8>: C6 C13 C10 HD9
> <TMS 8>: C3 C9 C5 HC4
> <TMS 8>: C5 C13 C9 HD8
> <TMS 8>: C15 C10 C13 C9
> <TMS 8>: C13 N1 C15 O1
> <TMS 8>: C16 C15 N1 H1
> total 102 improper torsions applied
> 32 improper torsions in old prep form
> Building H-Bond parameters.
> Parameter file was not saved.
>
> The problem is: I imported the residue prep file. However, how can I get the
> bond parameters about bonds connecting the residue and DNA? like o-P bond in
> the error message here? I will attach all the related files.
> Thank you very much!
>
> Best wishes!
> Yikan
>
>
> On Sun, Sep 6, 2009 at 8:30 PM, case <case_at_biomaps.rutgers.edu> wrote:
>
>> On Sat, Sep 05, 2009, Yikan Chen wrote:
>>
>> > Actually I do not know how to use frcmod file(I do not quite understand
>> the
>> > meaning of P70-71 of Ambertools manual), if I do what they said, there
>> will
>> > be:
>> > > mods=loadamberparams frcmod
>>
>> This is correct.
>>
>> > > saveamberparm TMS prmtop inpcrd
>> > Checking Unit.
>> > WARNING: The unperturbed charge of the unit: -1.000000 is not zero.
>> > -- ignoring the warning.
>> > Building topology.
>> > Building atom parameters.
>> > Building bond parameters.
>> > Building angle parameters.
>> > Building proper torsion parameters.
>> > Building improper torsion parameters.
>> > total 16 improper torsions applied
>> > Building H-Bond parameters.
>> > Not Marking per-residue atom chain types.
>> > Marking per-residue atom chain types.
>> > (Residues lacking connect0/connect1 -
>> > these don't have chain types marked:
>> > res total affected
>> > TMS 1
>> > )
>> > (no restraints)
>>
>> Looks like everything is working....dac

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber