Q: how to use MACS3 for ATAC seq with "himmratac" option?

[kbarrr]

Use case
My data consist of BAM paired-end files already filtered by duplicates, mit reads, etc. (99% alingment)

Describe the problem
I tried to use this command in bash

macs3 hmmratac -i M1_dedup_noM_sorted.bam -f BAMPE -g hs -n test -B -q 0.01

and got this error:
usage: macs3 [-h] [--version]
{callpeak,bdgpeakcall,bdgbroadcall,bdgcmp,bdgopt,cmbreps,bdgdiff,filterdup,predictd,pileup,randsample,refinepeak,callvar,hmmratac}
...
macs3: error: unrecognized arguments: -g hs -B -q 0.01

Additional context
macs2 works well, but i still want to learn MACS3.

Any help would be useful. Thank you.

[taoliu]

Hello @kbarrr the tutorial on using hmmratac command in the MACS3 (version 3.0) can be found here:

https://macs3-project.github.io/MACS/docs/hmmratac.html

The command line options are different from the callpeak command in MACS3.

Let us know if you find any issues when running hmmratac in MACS3. We are still optimizing it. @philippadoherty is currently working on reducing the memory usage for this command.

[taoliu]

@kbarrr Thanks for the information! However, the resolution of the cutoff analysis report is not good enough for me to judge where the cutoff should be -- each step is 10 according to the first column. I would guess the settings of '-u 20 -l 10 -c 2' (or -c 5) should work. We will release a new version soon to address 1) the resolution of cutoff analysis -- to make it with a step of 1 by default #642; 2) to reduce memory usage by a magnitude for large datasets #640 ; 3) to provide a simpler HMM model with Poisson Emission #635. If you can wait, please try the next release.