Which Statistics Should Be Applied To Determine Whether A Certain Motif Is Significant Among Amount Of Dna Sequences?

Question

[dustar1986]

Hi,

I have got 5000 DNA sequences. And I used MEME to find enriched motif among them.

It returned 134 sequences sharing a certain motif.

Which statistical test should I apply to determine whether such 134 out of 5000 is significant or not?

Thanks in advance.

[jack]

You can compute P-value for them.

[dustar1986]

Got it. Thanks a lot.

Answer 1

[dariober]

MEME should return the evalue for each motif it finds. I think the evalue is the statistics your are looking for. From this page:

The E-value is an estimate of the expected number of motifs with the given log likelihood ratio (or higher), and with the same width and number of occurrences, that one would find in a similarly sized set of random sequences. (In random sequences each position is independent with letters chosen according to the background letter frequencies.) Motifs with E-values larger than 0.01 (1e-2) are possibly just statistical artifacts, and not real motifs

So if the evalue for your motif is <<0.01, then 134/5000 sequences having your motif is a significant enrichment.

[karl.stamm]

This is the right answer. If you have set the background genome correctly, because the baseline for randomness looks quite different between mammals and fungi or virii...

[dustar1986]

Thanks a lot. This is exact what I want. I should have read the manual carefully.