This is an exciting and timely question. Over the last 2 years or so it has
become clear that CML is a very flexible language and that it is normally
possible to create a syntax that *can* hold a particular set of information.
However it generally not a good idea to try to enforce this in schemas. The
current approach is to use the "convention" attribute which means that a
subgroup of the community agrees on a particular semantics for a particular
syntax.
What I suggest is that the bluobelisk community discusses whether there is
an agreement on the components of a conformer data structure - what
information is mandatory and what is optional. To get the discussion going I
would suggest a simple approach like:
<cml xmlns="http://www.xml-cml.org/schema"; convention="cml:conformerList">
<molecule id="m1_1"...> contents... </molecule>
<molecule id="m1_2"...> contents... </molecule>
<molecule id="m1_3"...> contents... </molecule>
...
</cml>
The writers and readers would conform to this syntax. Note that there is an
implict requirement that all molecules have the same connTable and differ
only in coordinates. It is possible to cut out repeated info - e.g. bonds -
but I doubt it's worth it.
<cml> ...</cml> is a general container and can be nested so it's possible
to include many sets of conformers.
I can't see any immediate problem in this design, it's simple to understand
and parse. But it's easy to overlook things and it's important to get
community feedback.
On Mon, Aug 3, 2009 at 4:37 PM, David C. Lonie <[email protected]> wrote:
> Hi list,
>
> I'm interested in adding conformer support to the CML file format
> reader/writer in OpenBabel. Where can I find information regarding
> information as to how CML conformers should be written?
>
> Thanks,
>
> Dave
>
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus
> on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> cml-discuss mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/cml-discuss
>
--
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Blueobelisk-discuss mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/blueobelisk-discuss
