CLIP-seq analysis - RNA motifs

[adomingues]

Hi,

Is it possible to use gimmemotifs for the detection of RNA motifs from CLIP-seq derived locations? Off the top of my head, the main differences from ChIP-seq (DNA) motifs is that RNA motifs will be on a single strand, that to which the RNA-binding protein binds to, rather than both strands. With that in mind I have called binding site with a CLIP-seq specific tool, extracted the DNA sequencing from the binding strand, and used this command for motif discovery:

gimme motifs pred.rat.top1000.fa ./ -g ${GENOME}/Sequence/WholeGenomeFasta/genome.fa -N 10 --denovo --size 50 --analysis small --singlestrand

I am assuming that using --singlestrand only the sequences in the file will be use for motif prediction, and not those from both strands. Is this correct? --size 50 --analysis small are being used because there is a published motif for this particular protein and it is small - my goal is to partly recover the published motif from my data.

Am I missing nay other setting or doing this entirily the wrong way?

Cheers.

[simonvh]

Hi @adomingues

First, I have to say that I have never used gimme motifs in the way, nor am I familiar with others that have used it. That being said, it should work? Your assumptions are correct. The --singlestrand argument will make sure that the de novo motif prediction is only performed on the input sequence, not on the reverse complement. However, not all de novo motif tools support this, so it may vary.

[adomingues]

Thanks a lot for the input @simonvh.

However, not all de novo motif tools support this, so it may vary.

I will keep this in mind when looking at the results.