Warren A. Kibbe
Allyson’s note: I missed the beginning of this talk due to me participating in the press conference. Apologies.
Integrating clustering results is followed by final curation of DOLite terms, where a domain expert reviews the merged clusters. In summary, DOLite is a CV whereas the DO is an ontology. The purpose of this was to facilitate the funcational analysis based on a gene list. FunDO: website exploring genes using Functional Disease Ontology Annotations. You take the complete DO, and put in a typical gene list from a microarray study, and you get a network view of clustered genes. The same query using a gene list with DO and DOLite gets better clustering with DOLite (where better == more distinct clusters, greater number of clusters in the example we were shown – 2 clusters rather than 1).
You can also use GeneRIFs as a source (1000 genes with GeneRIFs annotations). You get slightly different answers depending on how you develop/annotate your gene list. Poorly-annotated genes, or a large % of genes with little or no exp literature will have few GeneRIFs.
Grouping ontology terms based on gene-to-ontology mapping provides a IC method for creating “Slims” from any type of ontology. They’ll do this with GO itself and see what their version of GOSlim looks like. Functional analysis based on DOLite provides much more concise and biologically-relevant results.
Please note that this post is merely my notes on the presentation. They are not guaranteed to be correct, and unless explicitly stated are not my opinions. They do not reflect the opinions of my employers. Any errors you can happily assume to be mine and no-one else’s. I’m happy to correct any errors you may spot – just let me know!