ERROR: Could not compute emission probabilities!

[tianxiahuihui]

Dear Sir:

Hello, My pureclip version is 1.3.1. I meet an error like this "ERROR: Emission probability became 0.0! This might be due to artifacts or outliers. ERROR: Could not compute emission probabilities! ". What is the problem? and how to solve it?

I found it showed this error when I used the bam that remove duplicate reads. Why this happened?

Thank you.

[skrakau]

Hi, which command did you use exactly to run PureCLIP or was it with default parameters? And for which type of data did this happen?

This error was reported recently already for some iCLIP2 data with considerably higher peaks compared to older protocols (causing decreased emission probabilities, for which the arithmetic precision was then not sufficient anymore).

Not sure if there is a ready to use solution. Maybe some parameter changes (can't say without knowing more about your specific case) or subsampling the data might help.

[tianxiahuihui]

Thank you. Maybe it is the problem of my bam file. My bam file is irclip-seeq bam file. Does the PureClip can be used in the irclip-seq data? irClip-seq methods article ( https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5477425/ ).

Thank you.