AMBER Archive (2009)

Subject: Re: [AMBER] different comformations to choose for RESP

From: FyD (
Date: Mon Jun 01 2009 - 09:32:21 CDT

Dear Xueqin Pang,
> I intend to get the partial charges with RESP after optimization
> with Gauss. However the output of Gauss for the lipid (dppc) differs
> a lot from the input  conformation (I intend to attach the file
> behind but failed).

You wrote: "The output of Gaussian for the lipid differs _a lot_ from
the input conformation". Here, what does mean "a lot" ?

And where does this input conformation come from ? from MD simulation
? is it an averaged MD snapshot ? was it optimized using a FF ?

> So I want to ask: which file should I use in RESP, the input (one of
> the lipid molecular in a equilibrated lipid membrane) or the output
> (optimization of the input HF/6-31G*)?

I would vote for using QM optimized geometry in MEP computation;
trying to understand why what you got from QM is so different from
what is this input conformation.

> What is more I get the partial charges of both the input and output
> confirmations of dppc with AM1-BCC, and the results is attached
> behind. Again I am wondering which one to use in my MD stimulation.

I would vote for using RESP charges. However, you could generate both
sets of charges, run your simulations and see why charge set your

regards, Francois

There are data in R.E.DD.B. about glycolipids you could have used but
they are protected up to 12-03-2010. See

AMBER mailing list