Hello all, I have used the Nextflow Nfcore/atacseq pipeline to analyze my data mostly with success. Upon attempting to run macs2 peak calling utilizing different parameters I discovered that my params.yaml file is not being correctly integrated into the pipeline. For example, I originally ran the pipeline with macs_FDR: 0.01 and then reran it again using macs_pvalue; 0.01. Even with this relaxed threshold I received the same number of peaks. Also after checking the narrow peak .xls file for both runs in the headers I found that the qvalue cut off was set to 0.05 which is the default. After checking the logs here is the actual command being run (no macs2 p value is specified so peak calling is run with defaults)
#!/bin/bash -euo pipefail
macs2 \
callpeak \
--keep-dup all --nomodel \
--gsize 2862089864 \
--format BAMPE \
--name MSC.mRp.clN \
--treatment MSC.mRp.clN.sorted.bam \
This is what my params file looks like
input : '/ATAC/samplesheet_nf_ATAC_downsample.csv'
outdir: '/results'
email: ch
multiqc_title: 'Osteo_chrombpnet'
seq_center: 'MedGenome'
read_length: 150
genome: 'GRCh38'
save_reference: TRUE
macs_pvalue: 0.01
narrow_peak: TRUE
And finally the command I used to run the pipeline
nextflow run nf-core/atacseq -params-file /scripts/params_chromBPnet.yaml -profile singularity
I have also added an image from the multiqc report that reveals the peak calling options that have been recorded.
I am not sure what could be causing this issue and if anyone has any information that could help regarding params files or any other configs that I might be overlooking that would be most helfpul.