On 19 Dec 2006, at 15:09, Paul Fisher wrote:
Dear Team,
I am currently working towards my PhD looking at a large-scale
systematic analysis of genotype-phenotype correlations and have been
using Taverna as the means of implementing this using workflows.
Over the past year I have managed to implement a novel means of
identifying candidate genes involved in the expression of a phenotype
and have obtained some significant biological results, which are being
put forward for being published at the moment in biology and
bioinformatics journals. These workflows utilised the BioMart Ensembl
processor available in Taverna.
However, since BioMart dramatically changed the schema, the
implementation of BioMart in Taverna can no longer run, which in turn
means that these workflows are not currently runnable. As you can
appreciate, this has had a disastrous knock-on effects for both my
own and for the Wellcome Trust Pathogen-Host project who wished to use
my workflows for processing their data. It has also meant that I am
not able to submit my papers to the proposed journals until I can
verify that the workflows are functional again and provide the same
means of data analysis. Being on the Taverna users list I have also
noticed a number of other researchers have had workflows fail because
of BioMart.
The Taverna BioMart developer (David Cc'd) has provided me with a
runnable version of BioMart 0.5, however, a number of attributes have
now disappeared that are vital to my methodology, including Cow-Human
Ensembl gene homologues. These issues are on the BioMart end and not
Taverna.
I am very grateful for the web service the BioMart team has provided
and the continual co-operation the BioMart developers have had with
the myGrid team, and hope to continue to use it for the remainder of
my PhD. However, as this is a large user group I think it would really
benefit both the myGrid developers and in turn the end users if some
means of email notification were provided on the Taverna mailing
lists, and likewise to subsequent high usage mailing lists to notify
people of any dramatic changes which may affect future research.
Hi Paul,
thanks for your input. The homologues did not disappear but simply have
moved
to Ensembl homology with the release of Ensembl v 42 as the Ensembl
compara
team deemed this to be better supported this way. As far as I know you
can now retrieve
much more information wrt to homologies through the new dataset.
Can you please see if you can adapt your workflows to talk to this
dataset instead?
If not we'll try about other means of improving this so your workflows
can be re-instated
and supported properly
a.
sincerely,
Paul Fisher.
------------------------------------------------------------------------
-------
Arek Kasprzyk
EMBL-European Bioinformatics Institute.
Wellcome Trust Genome Campus, Hinxton,
Cambridge CB10 1SD, UK.
Tel: +44-(0)1223-494606
Fax: +44-(0)1223-494468
------------------------------------------------------------------------
-------
