Identifying Motifs In A Geneset

Question

[Dataminer]

Hi!

I have a gene set of (100 genes), and I also have 2 transcription factors (say TF1 and TF2).

What I want to know from this data set is, how many of the genes in geneset have the motif for TF1 and how many have the motif for TF2 and how many have the consensus motif for both of them?

Any tools available for the same or any suggestions for the approach that can be adapted?

Thank you.

[Larry_Parnell]

The organism you're studying informs the choice of how much upstream sequence you take to analyze for the occurrence of motifs for TFs 1 and 2.

[Chris Evelo]

Any clues to what you would use for which species and why?

[Larry_Parnell]

It is a general rule of thumb, but I use genome length and gene density. Some knowledge of the literature helps, too, in determining how often near, middle distance and far upstream gene control elements have been identified.

[Dataminer]

In human.........

Answer 1

[Neilfws]

There are several ways into this problem. It would be nice to find a web server where you simply upload your nucleotide sequences and TF binding sites are returned. Unfortunately the Web is littered with examples of broken and disused tools for that purpose :)

Most people would start with the MEME suite, which has (limited) online tools and software to download and run locally. The FIMO program is a good start point. You may also want to download databases from JASPAR, which can be used with the MEME programs to search sequences.

You may also be able to construct a clever query using the UCSC table browser and/or Galaxy, to retrieve IDs for your genes and link them to one of the TF database tables, e.g. the ORegAnno track.