idr threshold
shinlin77 opened this issue · 1 comments
On my own data, I received this error
Exception: File is empty (rep5-pr1_vs_rep5-pr2.idr0.05.narrowPeak.gz). Help: No IDR peaks found. IDR threshold might be too stringent or replicates have very poor concordance.
Traceback (most recent call last):
File "/home/sl/.conda/envs/encode-atac-seq-pipeline/bin/encode_task_idr.py", line 213, in
main()
File "/home/sl/.conda/envs/encode-atac-seq-pipeline/bin/encode_task_idr.py", line 180, in main
'No IDR peaks found. IDR threshold might be too stringent '
File "/home/sl/.conda/envs/encode-atac-seq-pipeline/bin/encode_lib_common.py", line 218, in assert_file_not_empty
raise Exception('File is empty ({}). Help: {}'.format(f, help))
Exception: File is empty (rep5-pr1_vs_rep5-pr2.idr0.05.narrowPeak.gz). Help: No IDR peaks found. IDR threshold might be too stringent or replicates have very poor concordance.
How might I go about changing the IDR threshold? Alternatively, if there is poor concordance, is there a way to figure out which file is the problem? Thanks.
Next time please post pipeline's issues on the corresponding pipeline's git repo.
Please find a final (filtered) BAM file and count number of reads in there.
$ find -name *.nodup.bam
$ samtools flagstat FOUND_BAM
Please refer to this document about IDR threshold. The pipeline is just a wrapper to this software.
https://github.com/kundajelab/idr