search

philippdre / omniCLIP

omniCLIP is a CLIP-Seq peak caller
GNU General Public License v3.0
15 stars 9 forks source link

Description of output files #2

Open khembach opened 5 years ago

khembach commented 5 years ago

It would be great if you could describe the output files and especially the different columns in pred.txt. Is column pv the log10 of the p-value of the peak and what is the SiteScore? I assume that column TC is the number of T->C conversions, but are other diagnostic events, e.g. deletions, also reported? Thanks.

philippdre commented 5 years ago

Yes, pv the log10 of the p-value.

SiteScore is the log-likelihood ratio of the peak state versus the other states.

The TC column counts the number of T->C conversions.

khembach commented 5 years ago

Thank you! Does that mean that only the T->C conversions are reported but not deletions? I am asking, because I work with HITS-CLIP data and I would like to know the number of deletions per peak.

philippdre commented 5 years ago

Yes, it is still a legacy from working with PAR-CLIP data. I will try to change it in the next version.

khembach commented 5 years ago

Ok thanks, than I will try to find a way to compute the deletions.

dataMaster-Kris commented 5 years ago

Have the columns in the output files been described anywhere yet? It will be much appreciated if there was a proper documentation of how to interpret each of the columns in the tables (the above discussion helps with interpreting pv, SiteScore and TC but I am having to figure out the rest on my own). Are the p-values in pred.txt Bonferroni corrected? Also, what is the difference between pred.txt and pred.bed?