Visualizing the most likely ends of transcripts

Question

[pooryamb]

Hi all,

I have a distribution of 3' and 5' ending sites of a transcript. I want to visualize it like the following photo:

photo from "https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4748824/"

Do you know how can I do it? Is there a dedicated software package for these diagrams?

[Pierre Lindenbaum]

https://weblogo.berkeley.edu/logo.cgi ?

[pooryamb]

Thanks for your reply. I think in weblogo, the height of each letter in each position depends on its relative frequency compared to other alphabets in the same position and multiple letters might appear in each position. Here, I want the height of each letter would be proportional to the relative frequency of seeing a transcript ending in that position (compared to other positions). Because of the definition, always we would see just one letter in each position.

[LauferVA]

yeah im still with pierre though. did you download and run CircTail (the zipped file provided in the manuscript).

my guess is that they either modified weblogo, it is in circtail, or they wrote it custom. id check out circtail.zip. http://trypsNetDB.org/circTAILseqAnalyzer.zip which is linked in the manuscript. if its not there, either, the next two options id consider would be 1) contact the authors or 2) write it yourself.

However, because they explain how other figures were generated (e.g. Figure 3, weblogo), granted that they do not describe this process, it may mean they wrote a DoIt script or that they may have just used weblogo in a creative way, i.e. just fed it the probabilities calculated from the process you describe rather than from the normal process that you also describe xD