rbp_fimo_summarise.R - RBP density score doesn't normalise for motif length?

[SamBryce-Smith]

Hi Sid,

Was just trying to do some exploratory analysis to query RBP binding to some of my 3'UTRs of interest and thought your approach would be a good starting point.

In your graphic, the equation for your binding score suggests that it normalises motif counts for the length of the motif and the query region. When I check the script, it only seems to normalise by the length of the query region e.g. at line 25

Obviously a straightforward fix. Just checking if the code/score calculation is up-to-date with what you did for the manuscript?

By the way, props for sharing analysis code & documenting as well as you did (unfortunately still a rarity these days...), was pretty straightforward for me to reproduce your workflow :)

Cheers, Sam

[sid-sethi]

Hi Sam,

You are correct, we did not normalise for the motif length. As far as I remember, I think it gave us similar patterns since we were looking at an overall picture of all RBPs together, rather than comparing RBPs with each other. As you have spotted, the scoring equation in the graphic should be integrated in case you wish to take into account the motif length. Apologies for the confusion.

Thanks very much for the kind words. I am glad you found the documentation useful :)

Regards, Sid