Strand assignment problem for junctions

[shiro-kur]

Hi ! trackplot is a great tool. So convenient.

I found a strand problem.

I tried to perform $ trackplot
-e chr9:112296343-112335026
-r Homo_sapiens.GRCh38.101.chr.gtf.gz
--density bam.tsv
-o PTBP3.include.pdf
--dpi 300
--width 6
--height 1 --show-junction-num
--included-junctions chr9:112297917-112330441:-,chr9:112297917-112333470:-,chr9:112330475-112333470:- ,but I could not find any junctions.

Finally, I found the error for the strand assignment. Only with the options above, the all junctions were assigned to plus.

I would like you to improve the problem. Maybe the problem is in the options for library strandness.

for jxn_idx, jxn in enumerate(jxns_sorted_list): print("jxn_idx,jxn" ,jxn_idx,jxn ) ---output---- jxn_idx,jxn chr9:111641804-112514802:+ 0 jxn_idx,jxn chr9:111641804-112514447:+ 1 jxn_idx,jxn chr9:112268196-112333470:+ 2 jxn_idx,jxn chr9:112276014-112333470:+ 3 jxn_idx,jxn chr9:112297917-112333470:+ 4 jxn_idx,jxn chr9:112297917-112333103:+ 5 jxn_idx,jxn chr9:112297917-112332894:+ 6 jxn_idx,jxn chr9:112297917-112332784:+ 7 jxn_idx,jxn chr9:112297917-112330441:+ 8 jxn_idx,jxn chr9:112268196-112297841:+ 9 jxn_idx,jxn chr9:112268196-112297832:+ 10 jxn_idx,jxn chr9:112276014-112297841:+ 11 jxn_idx,jxn chr9:112276014-112297832:+ 12 jxn_idx,jxn chr9:112294511-112297832:+ 13 jxn_idx,jxn chr9:112330475-112333470:+ 14 jxn_idx,jxn chr9:112330475-112332894:+ 15

[ygidtu]

Hi shiro-kur: I have reviewed the code logic, and based on our example data, the functionality appears to be working correctly for filtering purposes. The --included-junctions parameter is specifically designed to retain junctions from the BAM file, and the filtering process can be observed in lines 279-280 of the trackplot/file/Bam.py file.

To assist with further debugging, could you please provide a detailed list of the junctions retained before and after the filtering process? This information will help us identify any potential issues and ensure the functionality is performing as expected.

Looking forward to your response.

Best regards, Yiming

[ygidtu]

Hi shiro-kur: We've updated our repository with a new version of the code (not yet officially released). We've modified the logic for judging --included_junctions and added more detailed debug logs, which might help you with your issue. You can consider pulling the latest code to test it out.

If you encounter the same problem, please run trackplot with the parameters --verbose and --logfile, and upload the generated logfile for us to assist in further debugging.

For example,

git clone https://github.com/ygidtu/trackplot trackplot
pip install -r requirements.txt

python main.py \
  -e chr9:112296343-112335026 \
  -r Homo_sapiens.GRCh38.101.chr.gtf.gz \
  --density bam.tsv \
  -o PTBP3.include.pdf \
  --dpi 300  \
  --width 6  \
  --height 1 --show-junction-num  \
  --included-junctions chr9:112297917-112330441:-,chr9:112297917-112333470:-,chr9:112330475-112333470:-  \
 --verbose --logfile trackplot.log

Looking forward to your response.

Best regards, Yiming