Hello, I am exploring the idea of creating a package to import pathways as BioPAX/SBPAX (Level 3) data to analyze various measurements. In particular, differential microarray measurements could be used to identify upstream pathway nodes that seem to play a critical role in explaining the observed differences. The basic idea is very simple: consider a pathway that contains reactions A -> X, B -> X and B -> Y. If measurements show an increase in X, but not in Y, this would suggest an increase in A. If, however, we see increases of both X and Y, then this would point to an increase in B. Analogous considerations apply to nodes upstream of A and B. Negative correlations (by inhibition or depletion) will also be included. Consider this applied to a large network and a large set of measurements, which requires statistical tools to identify the most relevant upstream nodes. There are people using similar methods on similar data relying on quite simple evaluation functions and turning it into a profitable business. We can start with a simple prototype that can be created and deployed as quickly as possible as proof of concept to find interested parties. If there is sufficient interest, a more sophisticated version can be built. Different implementation approaches are possible. It seems to be simple and efficient to use rjava and have the reaction network extracted from BioPAX/SBPAX by a Java package that uses OpenRDF Sesame Rio (i.e. a small fraction of Sesame dealing with RDF graph representation and I/O). Any comment or show of interest is greatly appreciated. Thanks! Take care Oliver -- Oliver Ruebenacker Bioinformatics Consultant (http://www.knowomics.com/wiki/Oliver_Ruebenacker) Knowomics, The Bioinformatics Network (http://www.knowomics.com) SBPAX: Turning Bio Knowledge into Math Models (http://www.sbpax.org)

this seems clever, although I could swear that an existing package handles graphical models of similar sorts. but if the structural priors can be automatically pulled in (however that might happen) and updated it would be cool as hell. On Thu, May 31, 2012 at 2:47 PM, Oliver Ruebenacker <curoli@gmail.com>wrote: > Hello, > > I am exploring the idea of creating a package to import pathways as > BioPAX/SBPAX (Level 3) data to analyze various measurements. In > particular, differential microarray measurements could be used to > identify upstream pathway nodes that seem to play a critical role in > explaining the observed differences. > > The basic idea is very simple: consider a pathway that contains > reactions A -> X, B -> X and B -> Y. If measurements show an increase > in X, but not in Y, this would suggest an increase in A. If, however, > we see increases of both X and Y, then this would point to an increase > in B. Analogous considerations apply to nodes upstream of A and B. > Negative correlations (by inhibition or depletion) will also be > included. Consider this applied to a large network and a large set of > measurements, which requires statistical tools to identify the most > relevant upstream nodes. > > There are people using similar methods on similar data relying on > quite simple evaluation functions and turning it into a profitable > business. > > We can start with a simple prototype that can be created and > deployed as quickly as possible as proof of concept to find interested > parties. If there is sufficient interest, a more sophisticated version > can be built. > > Different implementation approaches are possible. It seems to be > simple and efficient to use rjava and have the reaction network > extracted from BioPAX/SBPAX by a Java package that uses OpenRDF Sesame > Rio (i.e. a small fraction of Sesame dealing with RDF graph > representation and I/O). > > Any comment or show of interest is greatly appreciated. > > Thanks! > > Take care > Oliver > > -- > Oliver Ruebenacker > Bioinformatics Consultant ( > http://www.knowomics.com/wiki/Oliver_Ruebenacker) > Knowomics, The Bioinformatics Network (http://www.knowomics.com) > SBPAX: Turning Bio Knowledge into Math Models (http://www.sbpax.org) > > _______________________________________________ > Bioconductor mailing list > Bioconductor@r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > -- *A model is a lie that helps you see the truth.* * * Howard Skipper<http: cancerres.aacrjournals.org="" content="" 31="" 9="" 1173.full.pdf=""> [[alternative HTML version deleted]]