Discussion List Archives

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: CIF dicsussion on chemical structure

  • To: corecifchem@iucr.org
  • Subject: Re: CIF dicsussion on chemical structure
  • From: Peter Murray-Rust <pm286@cam.ac.uk>
  • Date: Wed, 06 Oct 2004 13:29:28 +0100
Thank you for asking me to comment on previous discussions and adding me to 
the mailing list. I note that you have mailed the essence of my comments 
yesterday to the list - thanks.

I have now read the past discussions so have a feel for what has informed 
the current draft. As I said I think it's very important that corecifchem 
is looking at this aspect. The discussions are particularly valuable in 
that we are writing software to deal with several aspects.

I include further mail from/to David and Brian today:

And I will be happy to help with the communal design. This is an important 
area and critical for what we want to do. And CIFs have a key role to play 
in the development of chemical informatics - Brian knows my views and 
aspirations. The point is that a CIF is, in principle, a formal validated 
publication of a chemical structure as well a crystallographic one. For 
whatever reason most crystallographers have not added the chemical 
information and so we have some ground to make up.

I think we can change this. It requires an OpenSource approach as otherwise 
we cannot control the direction. We are starting to get good tools for this 
which are recombinable. When peaking with John Davies recently he showed 
how he edited crystallography to chemistry on his Xray-machine. His system 
is not redistributable for technical and legal reasons but I think what we 
now have is. I shall discuss this with Brian next week. In simple terms:
- chemist gives crystallographer sample
- crystallographer solves sample
- software creates chemical structure diagram from crystallographic results
- crystallographer (and possibly chemist) edits structure to give best 
assumed chemistry.
- crystallography and chemistry are co-published.

I think this will work ca 90% for organic molecules. That is my immediate 
target. The main problems are, as we said:
- missing atoms
- symmetry
- disorder

I think the key thing is to come out with:
1:1 atom map for explicit non-disordered atoms
  precise count of non-scattering chemically important atoms
  precise count for disordered atoms
  complete connection table for chemistry

I am confident that CML can do this with minor modifications (mainly how to 
label symmetry and disorder). It also provides chemical editing for the 
structure. The result can then be written out as CIF or CML as required.

>When I have the other responses back I will again prepare a further 
>draft.  Following your advice I expect it will be a lot simpler and more 
>structured.  It is good to have a system that can be extended to be more 
>flexible, but I agree that in the first instance we need to focus on the 
>main task and not get carried away with what might in principle be possible.

Thanks - I think it's feasible.



Peter Murray-Rust
Unilever Centre for Molecular Informatics
Chemistry Department, Cambridge University
Lensfield Road, CAMBRIDGE, CB2 1EW, UK
Tel: +44-1223-763069

coreCIFchem mailing list

[Send comment to list secretary]
[Reply to list (subscribers only)]