Bug: Same output in all .bdg files when running --save-digested with HMMRATAC

[TeiturAK]

I have tried running macs3 hmmratac. I set the --save-digested flag because I want this for downstream analysis. I note that the bedgraph files contain the same signal in short-, mono-, di-, and tri- signals in all four bedgraph files in all regions.

I have analyzed public ATAC-seq data from https://pubmed.ncbi.nlm.nih.gov/33407087/. Data available at https://www.ncbi.nlm.nih.gov/Traces/study/?acc=SRP272433&o=acc_s%3Aa samples SRR12261697, SRR12261698, SRR12261699. These all have good periodicity in the insert size distribubion plots so I expect that these would work well.

I have attached a screenshot from a genome browser on what it looks like.

• OS: Linux
• Python version 3.9.12
• Numpy version 1.24.1
• MACS 3.0.0b1

I first tried this on newly generated data from our lab and got the same signal in all four output files.