5' promoter/alignment question

[bruce01]

I am looking at TFBS analysis for a colleague. He is interested in TRANSFAC but it being commercial I am interested in JASPAR. From Wasserman and Sandelin 2004 paper (doi:10.1038/nrg1315) they state "recurrent alignment of the 5' edges of ESTs and/or full-length cDNAs can be indicative of promoter locations".

While this may seem obvious, it was written 'pre-NGS' and as such I am wondering what the current consensus is on this? It is hard to search the literature for these sorts of issues and no good reviews have jumped out. To be more specific I was thinking about work stemming from this: is it worth a capture-seq experiment on differentially expressed genes to look for promoters, TFBS or other regulatory sequence from the 5' 'edge', or can you use the rather ad hoc assertion made above?

Thoughts appreciated.

[bb420]

Hi Bruce

I work for Biobase-International as a Field Application Scientist.

Despite (or because) being a commercial product, TRANSFAC has various
advantages over other products: its matrix library is much more comprehensive and more frequently updated , and is connected with a lot of detailed information, e.g. on the single binding sites on which the matrices are based. In addition, it contains a large number of processed high-throughput data, and various tools. Transfac has a very intuitive and a user friendly web interface. Also, our support team is always there to answer any questions.

I would also like to bring to your notice the recent Encode publication in Nature one of the papers compared Transfac and Jaspar and according to their results Transfac was more informative. You can have a look at the contribution of Transfac towards the Encode project here.
You can find more information on Transfac here.

Please feel free to reach me in case you need any further information.

Bhanu Bhargava
[email protected]

[ajthomas]

While the assertion is a good indication of promoter locations, it is certainly not conclusive. The only way to be sure is to map the transcription initiation site experimentally, and that's not always easy to do and can't be done in any sort of high-throughput fashion.

Another option to look at is the ECR Browser, assuming your species is one of the few whose genomes are available in that tool. It looks for conserved regions in genomes and can search for conserved and aligned TFBS. While still not conclusive proof that any given region is a functional promoter, evolutionary conservation of a region certainly adds more weight to the notion that it is.

[bruce01]

@AJThomas: Excellent, I had not previously encountered ECR browser and am fortunate that cow is a fairly well-studied beast. Many thanks for the suggestion.

@Bhanu: our licence for Transfac is pending!

[ajthomas]

Glad I could help. I also study the cow and the annotations in ECR Browser are pretty good for the cow. In case you didn't find it yet, the cow isn't one of the default species shown. Just click the blue "+" on the right side to add it.