Using pycistarget with bulk ATACseq data

[gamabunta313]

I am working on purified human b-cell populations. To call peaks I ran nextflow pipleline untill the generation of narrow peaks, then I used custom scripts for further processing.
I tried to apply pycistarget workflow on peaks to evaluate motif enrichment from my data.

All the packages were installed correctly and I did not observed errors.

To create the the cistarget motif database I ran the following script:

${create_cistarget_databases_dir}/create_cistarget_motif_databases.py 
-f ${region_fasta} 
-M ${clustertbuster_motifs} 
-m ${reference_database}/human_list_motifs.txt 
-o ${peaksname} 
-c $HOME/PROGRAMMI/anaconda3/bin/cbust 
-t 14 
-l 
-s 555

where:
human_list_motifs.txt= list of all transcription factor motifs found in human;
clustertbuster_motifs contains *.cb files download from "https://resources.aertslab.org/cistarget/motif_collections/v10nr_clust_public"
The I ran:

${create_cistarget_databases_dir}/convert_motifs_or_tracks_vs_regions_or_genes_scores_to_rankings_cistarget_dbs.py
    -i ${pycistarget_feather}/${peaksname}.motifs_vs_regions.scores.feather -s 555

At this point I run the following code in python shell:

ann = pr.read_bed(bedfile) 
data = cisTargetDatabase(rankings,region_sets = ann)
cisobj = cisTarget(
    name="Analysis",
    region_set = ann,
    species = 'homo_sapiens',
    auc_threshold = 0.0001,
    nes_threshold = 0.01,
    rank_threshold = 0.001,
    annotation_version = 'v10nr_clust',
    path_to_motif_annotations =annotations
)
cisobj.run_ctx(ctx_db = data)

where:
bedfile = generated by postprocessing and merging of selected peaks as pyranges object
ranking = file generated by convert_motifs_or_tracks_vs_regions_or_genes_scores_to_rankings_cistarget_dbs.py
annotations = .feather files downloaded from https://resources.aertslab.org/cistarget/databases/homo_sapiens/hg19/refseq_r45/mc9nr/gene_based

Everything ran successfully, but no targets were found.
I tried different cutoffs in terms of auc_threshold = [from 0.0001 to 0.001], nes_threshold = [from 0.1 to 3.0] and rank_threshold =[from 0.001 to 0.05].
None of there gave me 'significant' results.
Thanks in advance for all the support.