mfold parameter in MACS on Galaxy/ Galaxy Cistrome

Question

[nash.claire]

Hi everyone,

I'm afraid I am another newbie at ChIP-seq analysis (in general) and in galaxy but please help!

I have been doing a lot of reading that talks about the mfold parameter and is something that I'd like to experiment with. I've read that the MACS2 module automatically uses an mfold parameter of 5,50 as it's default (and this is what it says in my output when I run the MACS module with my alignment data).

However, I'd like to change this parameter but unless I'm being really stupid or really not understanding something (entirely possible), I can't find any options when running MACS2 with Cistrome to change this parameter? I've also tried going to using MACS in the usual public galaxy server but they only seem to offer MACS 1.0.1 and although this has a parameter for mfold, you can't seem to specify the lower and upper limits (like 10,30) it only takes one number with a default of 32.

Can anyone explain why I can't seem to find any MACS version on Galaxy/Galaxy Cistrome where I can set a parameter of 10,30 or 8,30 etc for mfold?? Or point it out to me if it's obvious and I'm missing it?

I look forward to your help!

Answer 1

[Ian]

My only suggestions involve you having access to a Linux OS.

You could set up your own version if you have access to a Linux machine: (virtualenv - running Python based software in isolation). This was made by me, be sure to reads the useful comments as well.

Or create a private instance of GALAXY: (https://wiki.galaxyproject.org/Admin/GetGalaxy). The are a couple of MACS2 tools in the tool shed that can be added, including the one we use (https://toolshed.g2.bx.psu.edu/view/pjbriggs/macs21/).

[nash.claire]

Thanks for getting back to me Ian. The thought of Linux is pretty scary at least for the time being!! I'll look into adding MACS2 tools for Galalxy. I've recently gotten access to my university instance of Galaxy so I'll try and add it on there. In the meantime I've just run with the default settings on Cistrome with MACS2 but the results don't seem to correlate well with a separate analysis of the same data that was run by someone else on MACS 1.4.2. I'm wondering if this is due to the mfold parameter or not but without being able to adjust it, it will be hard to figure out!!

Definitely need to learn R.