Re: [charmm-bbs] H-bond parameters (2)
- From: Don Gregory <dgregory |-at-| msi.com>
- Subject: Re: [charmm-bbs] H-bond parameters (2)
- Date: Tue, 12 Oct 1999 12:18:24 -0700
I'm glad you asked these two questions, because they point out some
important aspects I was trying to make in my earlier response.
Empirical force fields can *only* calculate those energy term they are
parameterized for, so if you are using a force-field that wasn't
parameterized using an explicit H-Bond term, it would be an inappropriate
question to ask "what is the H-Bond component".
To ask this question, you'd have to use a force-field that was
parameterized using explicit H-Bonding terms (which would no doubt
have other parameters different [e.g. electrostatics and vdW])
You can 'approximate' the strength of the H-Bonds by 'turning them on'
(i.e. incr. cuthb) but these numbers should *not* be relied upon too
greatly, since, again, those values were not used in the force-field
optimization. They are probably hold-overs from earlier optimizations
when they were included explicitly.
As for 'what do you do when there are no H-Bond terms in the force-field'
(even if you are using a field that has explicit H-Bond terms); the answer
is the same as any 'missing parameter' problem, i.e. you'd need to search
the literature to see if anyone has optimized such a parameter *using* the
same force-field you are using.
The script I provided still works with a Cys present, but you're right,
since there are not terms to calculate, one cannot even analyze for
H-Bond geometries (which is the recommended use of the H-Bond
commands I gave; also Lennart Nilsson reminded us of the
"coor hbond" command which helps with analysis w/o the use of
the force-field parameters).
At 04:25 PM 10/11/1999 +0100, you wrote:
>Thank you all for your help regarding my recent question on H-bond
>parameters. Following on from everything that has been said I would like
>to ask two more questions:
>(1) If hydrogen bonding is implicit in electrostatics and vdW terms, how
>do you analyse the energy term to find the contribution to energy which
>is due to hydrogen bonding? I assume this is a valid occasion for using
>(2) If you determine the energy due to hydrogen bonds using HBONDS CUTHB
>4.5 (for example), how do you account for sulfur containing systems,
>particularly cysteine residues, when sulfur is not parameterised in the
>HBOND section of the parameter file?
>If you run the script that Don Gregory sent to the list, but replace one
>of the residues with CYS this adapted script falls over because there
>are no hydrogen bond parameters.
>Thanks again for your help so far. Look forward to hearing from you
Dr. Don Gregory (dgregory |-at-| msi.com)
Asia Pacific Science Liaison
Molecular Simulations Inc.
9685 Scranton Rd.
San Diego, CA 92121
(619) 799-5331 http://www.msi.com