Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

New Version of the Ontologizer: Webstart, Data Exploration and Visualization, and Parent-Child Analysis

Peter Robinson peter.robinson at charite.de
Tue Feb 13 04:54:05 PST 2007


Hi all,

we would like to announce version 2 of the Ontologizer, which was 
launched in 2004 as a Java desktop application.

The Ontologizer now is a Java Webstart application that can be started 
directly from our webpage:

http://www.charite.de/ch/medgen/ontologizer/

The Ontologizer has a number of new features that may be interesting for 
members of this list:

1) Statistical analysis of term overrepresentation for 1 or more study 
sets. The results are displayed both in table form, whereby
detailed information about each term (definition, parents/children of 
the term, and a list of all annotated genes) as well as a GraphViz based 
graphical display of the significant terms are provided. You can jump 
directly to parent or children terms in order to explore datasets.

2) Implementation of a new procedure for statistical analysis of term 
overrepresentation, which we term "parent-child analysis". We have shown 
that
the standard approach using the Fisher exact test/hypergeometric 
distribution for each GO term leads to statistical artefacts because 
each term is considered in isolation. GO annotations inherited from more 
specific descendant terms can result in certain types of false-positives 
with potentially misleading biological interpretation.
We have found that by measuring overrepresentation of GO terms in the 
context of annotations to the parents of the terms, these dependencies 
can be greatly reduced,  thereby avoiding  producing false-positive 
results owing to this phenomenon.

A preliminary writeup of this algorithm was presented at RECOMB 2006 and 
was published in Springer's Lecture Notes in Computer Science series
(Grossmann, S., Bauer, S., Robinson, P. N. & Vingron, M. An improved
statistic for detecting over-representated gene ontology annotations in 
gene sets. In Research in Computational Molecular Biology: 10th Annual 
International Conference, RECOMB 2006, Venice, Italy, April 2-5, 2006, 
vol. 3909 of Lecture Notes in Computer Science, 85–98 (2006).). A fuller 
presentation of our method is now under review.

3)  Users can use their own gene annotation files to perform analysis. 
Routines for using affymetrix annotation files directly are now being 
developed.

Take care,

Sebastian Bauer, Steffen Grossmann, Martin Vingron, Peter N. Robinson


-- 
Dr. med. Peter N. Robinson, MSc.
Institut für Medizinische Genetik
Universitätsklinikum Charite
Humboldt-Universität
Augustenburger Platz 1
13353 Berlin
Germany
voice: 49-30-450569124
fax:   49-30-450569915
email: peter.robinson at charite.de
http://www.charite.de/ch/medgen/robinson


--
This message is from the GOFriends moderated mailing list.  A list of public
announcements and discussion of the Gene Ontology (GO) project.
Problems with the list?           E-mail: owner-gofriends at geneontology.org
Subscribing   send   "subscribe"   to   gofriends-request at geneontology.org
Unsubscribing send   "unsubscribe"  to  gofriends-request at geneontology.org
Web:          http://www.geneontology.org/



More information about the go-friends mailing list