I have the following problem: When I select the option --coverage-search in TopHat2, the junction file in the output directory is empty.
Using "--coverage-search" makes sense here, because I want to check for novel junctions in some old RNA-Seq data (36bp single-end and 38bp paired-end), so using segment-search is not a perfect option (splitting into 18bp segments will likely produce too many ambigous alignemnts).
The junction-files are empty both whith and without using a transcriptome annotation (human).
However, when I turn in the segmented search (by setting --segment-length to 18 or 19) together with --coverage-search, TopHat2 seems to return junctions from the coverage-search, because more junctions are returned using both options, than when only the segmented search is used.
The current TopHat-version installed on our cluster is "TopHat v2.0.8b"
Can anyone tell me whether this is a bug in TopHat2, or if I'm doing anything wrong?
Morten
Using "--coverage-search" makes sense here, because I want to check for novel junctions in some old RNA-Seq data (36bp single-end and 38bp paired-end), so using segment-search is not a perfect option (splitting into 18bp segments will likely produce too many ambigous alignemnts).
The junction-files are empty both whith and without using a transcriptome annotation (human).
However, when I turn in the segmented search (by setting --segment-length to 18 or 19) together with --coverage-search, TopHat2 seems to return junctions from the coverage-search, because more junctions are returned using both options, than when only the segmented search is used.
The current TopHat-version installed on our cluster is "TopHat v2.0.8b"
Can anyone tell me whether this is a bug in TopHat2, or if I'm doing anything wrong?
Morten