Analysis for Unknwon motifs obtained from Meme or Dream.

Question

[SP]

Dear All,

My question is regarding motifs which are found significant by Meme or Dream's de-novo search. What are the different approaches we can employ (biological or informatical) to further investigate these motifs?

Thanks, Peeyush

[Fidel]

Did you try to find out if similar motif's are already known? Even if you do de-novo search, chances are that the motif is already known.

You can also check the specificity of the binding proteins towards the discovered domain. Most likely, a single domain is not found for 100% of the regions. You can investigate why this happens.

[SP]

I have used Meme-chip. so motifs were already compared to Uniprobe mouse and JASPAR db. After centrimo analysis I found that some of these unknown motifs were listed at top ranks.

Could you please tell me how to "also check the specificity of the binding proteins towards the discovered domain"??

Thanks

[TriS]

if you have the different variants of that motif, clone them, put them on a luciferase vector and see if you expect that certain conditions activate the luciferase transcription. that would mean that those conditions allow certain TF to bind the region you identified and by having the different variants you will know the specifically which sequence gives you what.

for "certain condition" that has to be something related to your work...i.e. if you study TF Xyz then you can overexpress Xyz and see if that changes luciferase signal, or if Xyz knock down does the opposite. then, after having those data you can check, by either IP or by parsing databases of PPI (Protein Protein Interactions) which proteins directly interact with Xyz, which functions are they involved with ...etc...etc...etc....just tossing some ideas while having coffee ;)

hope this helps!

[SP]

Thanks TriS, this looks interesting.