I have already figured out a workaround to this issue, so it is not urgent, but I wanted to bring it to your attention in case others might be experiencing the same problem.
I was experiencing a consistent issue running the association pipeline in which the outputs would indicate an alarmingly low barcoding rate across all my libraries. I starting running some of the association code line by line and found that on line 57 of the nf_ori_map_barcodes.py script where it reads in the barcode fastq file, it was truncating the file extremely early (reading in <1% of the reads). The input fastq files I was using were straight from the Illumina BCL Convert pipeline. I did a bit of searching and found that this is a known issue with pysam < 0.16.0 as documented here.
The problem can be avoided by simply gunzipping the barcode fastq file (or gunzipping and regzipping). Perhaps you may want to hard code the gunzipping of the input fastq and/or use pysam >= 0.16.0 if it wouldn't break anything.
Happy to share a sample file if it would be useful in troubleshooting, but figured since it was already a documented issue with pysam < 0.16.0, it should be a pretty straightforward fix.
Hello,
I have already figured out a workaround to this issue, so it is not urgent, but I wanted to bring it to your attention in case others might be experiencing the same problem.
I was experiencing a consistent issue running the association pipeline in which the outputs would indicate an alarmingly low barcoding rate across all my libraries. I starting running some of the association code line by line and found that on line 57 of the nf_ori_map_barcodes.py script where it reads in the barcode fastq file, it was truncating the file extremely early (reading in <1% of the reads). The input fastq files I was using were straight from the Illumina BCL Convert pipeline. I did a bit of searching and found that this is a known issue with pysam < 0.16.0 as documented here.
The problem can be avoided by simply gunzipping the barcode fastq file (or gunzipping and regzipping). Perhaps you may want to hard code the gunzipping of the input fastq and/or use pysam >= 0.16.0 if it wouldn't break anything.
Happy to share a sample file if it would be useful in troubleshooting, but figured since it was already a documented issue with pysam < 0.16.0, it should be a pretty straightforward fix.
Best, Ryan