Hi Peter (and new Quixote project coworkers):
Good to hear from you Peter and to connect with you again, after a too 
long gap of some years :)! If only I still had our scribbled-on lunch 
napkins from 1995, all would be clear :)!

I also think there is a great current opportunity to connect some of 
your work, CML, CDK, Blue Obelisk initiatives, Bioclipse etc. with the 
new OpenTox initiative and we have the possibilities to make "quite 
startling progress" in coming months.  But we must also critically link 
OpenTox with pathways, systems biology and bioinformatics too, at least 
for supporting the growing integrated service and anaylsis needs of 
advancing the field of predictive toxicology.

The questions we have to answer and whose answering we need to support 
go something along the lines of
(http://opentox.org/home/documents/presentations/pathprot3presoct2010/view) 
:

/What are the directions for Integrated Analysis in Predictive 
Toxicology that extend current strategies, methods and infrastructure to 
better include development and support of a combination of statistical, 
mechanistic and pathways based approaches to optimise progress in the 
quality and acceptance of alternative testing methods?

How do we best leverage current knowledge and methods with regards to 
biological pathway analysis to design improved approaches to predictive 
toxicology that increase our ability to characterise the potential of 
chemicals to cause adverse human health effects and including an 
understanding of mode of action, mechanisms involved in the mode of 
action and the interaction of biological entities, pathways and networks 
in the perturbations introduced by the chemicals?/

/...and what variables need to be included in the interoperable service 
support design?/: Many including -
Contexts for: A. Pathway Known B. Pathway Unknown and Rational 
Hypothesis is formed C.Pathway Unknown and no Rational Hypothesis 
currently formed
and
Variables for:
- Most generally - any relevant interacting influencing variable in the 
complete environment of the planet
- Chemicals, genes, proteins, modifications to any of previous, 
metabolites (and their interactions)
- Species, Sex, Age, Diet, Organ(s) & Systems Interactions, Drugs, 
Genetics, Healthcare record, In vitro vs in vivo covariables
- Organ, Organelle, Tissue, Cell -- Cell Combinations, Metabolism, Cell 
(Type, differentiation), Sub cellular compartments e.g., mitochondria, 
Concentration, Time, Dosage (amount, frequency), Route of 
Administration, Reversibility of Toxicity effects, System interactions 
(local, non-local) .....other factors:)....

So we really have a Grand Interoperability Challenge almost demanding 
the benefits of a consensus-based Open & Public 
Standards/Ontology/Dictionary/Source/Data/Access approaches.  (All the 
good stuff Peter has been driving for in cheminfo for many years!!)

We ("OpenTox & Partners") are organising some workshop activity near 
Cambridge at EBI 15-17 November to discuss interoperability, data 
exchange standards, public ontologies, and a collaborative ontology 
development roadmap for the field of predictive toxicology.  In addition 
to good science & engineering, there is also an increasingly strong 
business case driver, so we also have the collaboration and involvement 
of the EBI Industry Forum and Pistoia Alliance.  We will also 
communicate outcomes aggressively with policy makers and program 
designers. Quo Vadis R&D progress and success in such scientific 
challenges without community, flexible infrastructure and interoperability?

Participation in the workshop at EBI involves no registration fee but 
there are costs and places are limited - please contact me if you have 
an interest in contributing to what should be a significant landmark 
event for this field's progress!

Barry

Am 25.10.2010 13:51, schrieb Peter Murray-Rust:
> On Mon, Oct 25, 2010 at 11:47 AM, Egon Willighagen<
> egon.willighagen at gmail.com>  wrote:
>
>> Nina,
>>
>> On Mon, Oct 25, 2010 at 12:28 PM, Peter Murray-Rust<pm286 at cam.ac.uk>
>> wrote:
>>> On Mon, Oct 25, 2010 at 10:33 AM, Nina Jeliazkova<
>> jeliazkova.nina at gmail.com>  wrote:
>>>> For gaussian files it should be feasible as well, does CDK reads
>> gaussian files ?
>>> JUMBO does and the transofrmation is designed to be lossless by using a
>> dictionary structure.
>>
>> CDK has only a minimal Gaussian readers, that extracts 3D coordinates,
>> and might actually do vibrations, but nothing much else.
>>
>> I think what we will want to do, is extend the CDK CMLReader to
>> understand the CMLComp conventions...
>
> That's exactly the right way to go. A typical CMLComp document (from a log
> file) might include modules such as init; optimisation, final coords,
> Hessian, Frequencies, NMR calculations.
>
>
>