https://nf-co.re logo
Join Slack
Powered by
Hello, we have some questions about the MACSima -&...
# mcmicro
a
Hello, we have some questions about the MACSima -> MCMICRO integration. It looks like MACSima ran two scans (SCN-001 and SCN-002) within each cycle, and each of them had two exposures. In addition, 3_Scan2 contains what looks like autofluorescence corrections for SCN-002. I assume these are combined to do the background correction with -S and -B. However, there are also two exposures for every marker x scan (high and low). We think this is for the dynamic range correction. macsima2mc is separating is exposure as a separate folder (-exp1 and -exp2), instead of doing the HDR correction. This results in two final full analysis for each "exposure channel". Is this behavior expected?
j
Hi Alejandro! The Schapiro Lab, who developed the Macsima-mcmicro workflow, hasn't been around here much lately. I don't have any experience working with this data, we just accepted their changes into the old mcmicro (version 1). They were planning to bring that work over to nf-core mcmicro (version 2) but I'm not sure where it stands right now.
I see on the macsima2mc github page that it had a new release just last week, so they're definitely still working on it.
@Victor Perez Should we reconnect on the macsima input support here?
a
That makes sense*, thank you for the info. We are working on implementing these two steps on our HPC to complement some workstations. It works really well, but this exposure channel difference is confusing us. Here is our documentation: https://github.com/StonyBrookRCI/MACSIMA_MCMICRO/tree/main Also, we are working from the RawData, but MACSima now makes the PreprocessedData on the fly and registers it.
j
You should be able to use the macsima2mc output with nf-core/mcmicro too, you'll just need to build your sample sheet csv
and markers csv with the relevant settings to enable background subtraction using the autofluorescence scans
(very similar to mcmicro 1 markers csv)
We don't have a Macsima instrument here in the LSP where the main mcmicro development is happening, so I can't help much with your specific questions about the multiple exposures for the AF scan.
a
Ok, so it sounds then like a module for macsima2mc is on the works to fix this. I will ask the other developers, I think the missing step that MACS IQview does is some type of HDR across the two exposures
And that the nf-core/mcmicro fully functional with the current dev version. That's also why I used the mcmicro v1
j
mcmicro v2 is very close to ready. you should be able to use the dev branch already.
🙌 1
Are you using 
--illumination_correction
in macsima2mc or are you running illiumination correction via mcmicro? Note that you only need one or the other.
a
Only on macsima2mc
j
OK. In mcmicro v2 we upgraded to the same 
basicpy
tool that macsima2mc is using. In case that is helpful to know.
It may or may not be easier to have that heavier computation running as part of mcmicro, instead of in the staging step.
a
Great, I will try then doing that. It is probably way better than what I am doing now, which is running job arrays and calling nextflow on each separate ROI
j
Ah! Yeah we now run all samples through one nextflow job
a
Wonderful
Thank you so much for your help! The main problem I encountered is just making the input recognized by MCMICRO. I will do this, and see if it all goes well
j
Once we do release v2 officially, I'd be happy to sit down with you and figure out how to make it work
a
Thank you! That would be super helpful
c
Hello, about the Macsima staging, Victor is not very active in Slack, but we are working on updating the module, and we can then add it to mcmicro?
👍 1
9 Views
Open in Slack
PreviousNext
Powered by