Understanding PlotAggregate output

[RolantusdataExp]

Dear devs,
Thanks for the great tool!

I have now run through the standard workflow, and reached the point where I want to visualise the results.. I have tried PlotAggregate, and reached the following results

CEBPA_footprint_comparison_all.pdf

NR3C1_footprint_comparison_all.pdf

What I am uncertain about is the interpretation of the shift in y-axis. In the first plot (CEBPA), the two plots seems to overlay quite well, but the second (NR3C1) the blue graph is shifted on the y-axis - making the interpretation of the footprint a little odd. Do you have a plausible cause of this, and is it a issue?

I have used the default:
$ TOBIAS PlotAggregate --TFBS test_data/BATF_all.bed --signals test_data/Bcell_corrected.bw test_data/Tcell_corrected.bw --output BATFJUN_footprint_comparison_all.pdf --share_y both --plot_boundaries --signal-on-x

[hschult]

Hi @RolantusdataExp,
Thanks for your kind words.

It could be either a normalization artifact, e.g., due to a different background used during calculation, or gWA has more cutsites around NR3C1 binding sites. It's hard to say without knowing how you calculated the bigwig files.

I suggest using our snakemake pipeline, which automatically runs the TOBIAS modules in order and takes care of file handling and parameters. Or if you want more details, you could look into our resources. Our wiki and the Supplementary of our related paper might be a good start.

Cheers,
Hendrik

[RolantusdataExp]

Hi Hendrik,

Thanks for the rapid response!

I have used the standard workflow suggested in your wiki. I have used the same peakset for the analysis. One thing I am suspecting is whether the difference in the number of reads between the conditions is too large. gWA has around 60% more reads than iWA - is that something I should be taking into account?

Best regards,
Peter

[hschult]

Hi Peter,

ATACorrect normalizes for read count. I suspect the issue is elsewhere.

[RolantusdataExp]

Hm, that makes it more tricky.. Here are​ all the commands I have run (simplified just to show 1 example):

ATACorrect!
TOBIAS ATACorrect --bam /work/CD_merged.bam --genome /work/mm39/mm39.fa --peaks /work/AllConditions_merged_narrow.bed --blacklist /work/ mm39.blacklist.v2.bed --outdir /work/Footprinting_iWAT/ATACorrect --cores 8

ScoreBigWig
TOBIAS FootprintScores --signal /work/CD_merged_corrected.bw --regions /work/AllConditions_merged_narrow.bed --output /work/CD_gWA_footprints.bw --cores 8

BINDetect
TOBIAS BINDetect --motifs /work/JASPAR2024_CORE_vertebrates_non-redundant_pfms_jaspar.jaspar --signals ScoreBigWig/CD_gWA_footprints.bw ScoreBigWig/CD_iWA_footprints.bw --genome /work/mm39.fa --peaks /work/merged_peaks_annotated.bed --peak_header /work/merged_peaks_annotated_header.txt --outdir BINDetect_CD --cond_names CD_gWA CD_iWA --core 4

Best,
Peter