NWChem/Quixote at EBI

[Peter Murray-Rust]

I am now collaborating with Christoph Steinbeck's group at the European Bioinformatics Institute on computing properties of molecules in the (human and other) metabolome. We have decided on using NWChem as the central engine and the JUMBO Parsers and converters to carry out file transformations. Today Merche (copied) and I installed and ran NWChem. The next phase is to compute chemical shifts for C,H and N in organic compounds.

So I/we expect to be going into a phase of activity in hacking NWChem input and output and any advice would be welcome. This is a medium throughput activity where we may compute several thousand molecules, probably with conformations and solvents.

P.

--
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069

[DeJong, Wibe A]

Hi Peter,

 

I should be able to help you to get the NMR inputs going in NWChem, and provide you insight into the output data you need.

 

Bert

[Peter Murray-Rust]

On Thu, Aug 11, 2011 at 10:59 PM, DeJong, Wibe A <Wibe....@pnnl.gov> wrote:

Hi Peter,

 

I should be able to help you to get the NMR inputs going in NWChem, and provide you insight into the output data you need.

Wonderful. I have already written template parsers for the an/isotropic  tensors so it should be straightforward. We are currently using QA/prop_ch3f.nw as a typical input. At some stage we need to :
* calculate the TMS as a reference and write a tool for the subtraction/offset
* calculate the lowest energy conformers
* model solvent.

P.

[Peter Murray-Rust]

On Fri, Aug 12, 2011 at 11:01 AM, Christoph Steinbeck <stei...@ebi.ac.uk> wrote:

This is brilliant. Thanks for the update. Glad to see this ball rolling.

Luis, Merche and I are pressing on. The BRMD data is variable and I'd like to discuss it in some detail with you - there are spectra of good and bad quality and some are truncated. In some cases  only some of the peaks are reported, etc. So I am not very convinced about immediate automation.

Merche and I will try to compute shifts for caffeine today...

P.

Cheers,

Chris

--
Dr. Christoph Steinbeck
Head of Chemoinformatics and Metabolism
European Bioinformatics Institute (EBI)
Wellcome Trust Genome Campus
Hinxton, Cambridge CB10 1SD UK
Phone +44 1223 49 2640


What is man but that lofty spirit - that sense of enterprise.
... Kirk, "I, Mudd," stardate 4513.3..

[Peter Murray-Rust]

On Friday we ran NWChem to produce isotropic shifts and by Mojnday should have enought info to see if we are on the right track (we are using alanine, TMS and caffeine to start with). Not sure yet how to extract the couplings from NWChem and create peak envelopes. JUMBO/Euclid has some routines for Gaussian smearing.

[DeJong, Wibe A]

Peter,

 

Don’t know what the CML translation looks like, but in NWChem the shielding output looks like:

 

Atom:   2    F

  Diamagnetic

    < Three lines of tensor >

 

  Paramagnetic

   < Three lines of tensor >

 

  Total Shielding Tensor

  < Three lines of tensor?

 

       isotropic = 457.3230

    anisotropy= 37.1539

 

What you’re interested in is the Atom and the isotropic shielding. Combining the TMS shielding with the alanine or caffeine will give the chemical shift (TMS-compound).

 

Bert

[Peter Murray-Rust]

Thanks,

On Wed, Aug 17, 2011 at 12:05 AM, DeJong, Wibe A <Wibe....@pnnl.gov> wrote:

Peter,

 

Don’t know what the CML translation looks like, but in NWChem the shielding output looks like:

We have parsed it into 3*3 matrixes and scalars
 

 

What you’re interested in is the Atom and the isotropic shielding. Combining the TMS shielding with the alanine or caffeine will give the chemical shift (TMS-compound).

Yes, we are in the process of checking against experiment. Will let you know how it goes.

 

Bert

 

 

 

From: quixot...@googlegroups.com [mailto:quixot...@googlegroups.com] On Behalf Of Peter Murray-Rust
Sent: Saturday, August 13, 2011 5:04 PM
To: Christoph Steinbeck
Cc: quixot...@googlegroups.com; mer...@ebi.ac.uk; Luis Filipe de Figueiredo; Mark Williamson
Subject: Re: [quixote-qcdb] NWChem/Quixote at EBI

 

On Friday we ran NWChem to produce isotropic shifts and by Mojnday should have enought info to see if we are on the right track (we are using alanine, TMS and caffeine to start with). Not sure yet how to extract the couplings from NWChem and create peak envelopes. JUMBO/Euclid has some routines for Gaussian smearing.

P.




--
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069