How to trace back methylated read IDs

[gsukrit]

Hi team,

I wish to trace back methylated read IDs from my data. However, after filtering >=0.9 probability_modified values in the file "data.site_proba", I have none of the transcript_id with >=0.9 values in the probability_modified column of the file "data.indiv_proba". I am slightly unclear on how this could be possible.
For example,

The gene has probability_modified of 0.91676724

The same gene when traced back in the data.indiv_proba file, at the same transcript position (1016), has no values >=0.9 in the probability_modified column:

Can someone explain this discrepancy? Also, kindly explain the right way to trace back the methylated read IDs (with >0.9 probability, as suggested by authors) corresponding to the methylated site with probability_modified of >=0.9

Thanks for the help!!

[yuukiiwa]

Hi @gsukrit,

The site-level probability_modified represents the probability that a site contains at least one modified read while mod_ratio is the proportion of reads at the sites above a predefined read threshold for a read to be considered modified.

Thanks!

Best wishes,
Yuk Kei