Motif Seq In Enriched Genes

Question

[kanwarjag]

It may appear to be very simple question. but I got lost. I ran and analyzed a Chipseq experiment. I identified binding motif for a particular protein of my interest. (which is around 7-8 letters). I scan the motif for presence with my list of enriched genes. I got a list of coordinates (100bp) theroetically which may have that binding motif. I want to see/ confirm at which position of enriched gene (it will be promoter) the motif seq is present. Is there any good way of doing in a set of genes say like 100 or so. Thanks

Answer 1

[Giovanni M Dall'Olio]

You may try with BEDOPS or BEDtools.

Put the coordinates of motif matches as a BED file:

chr1 1000 1008 motif1_match1
chr2 3000 3008 motif1_match2
chr3 4000 4008 motif2_match1
chr3 1200 1208 motif1_match3

Put the coordinates of your genes as a BED file, too:

chr1 1000 12008 gene1
chr2 1000 4021 gene2
chr3 8000 12008 gene3
chr3 1200 1208 gene4

If you want to include an upstream/downstream interval to these coordinates (to include the promoter and regulation regions), you can do it with gawk:

gawk "{print $1, $2-1000, $3+1000, $4}' mygenes.bed > mygenes_1000flank.bed

Then, you can use BEDOPS or BEDtools to get the intersection of these two files:

bedops --intersect motifs.bed mygenes_1000flank.bed

[kanwarjag]

One of the problem which I encounter is Chip-seq motif is 7-8 letters and when we intersect this way bed file we apply some sort of condition that 50% (0.05) or some thing that sort match. Will not be that if we have found motif and try to find out in the list of genes it should be close to 100% matching. Please correct me If I am off the track Thanks