CATALYST icon indicating copy to clipboard operation
CATALYST copied to clipboard

Integrating with Spectre Normalize for Batch Correction

Open ramziabb opened this issue 4 years ago • 5 comments

Has anyone integrated Spectre Normalize for batch correction (https://wiki.centenary.org.au/display/SPECTRE/Discovery+workflow+with+batch+alignment+using+CytoNorm) then the CATALYST workflow? I am working to doing so for both cyTOF and Flow data. Have gotten it working for the most part, now just sorting out if my transformation levels are correct and them focusing on analyzing results.

But I just wanted to reach out and see if others are using these workflows together.

Right now I am doing Spectre Normalize batch correction then writing out FCS files with:

write.files(cell.dat, file.prefix = "", divide.by = sample.col, write.csv = FALSE, write.fcs = TRUE)

I am then reading in the fcs files as my inputs into CATALYST. It works and so I don't really feel strongly about converting directly in the code.

Anyone has used these two workflows together? Has it worked for you?

ramziabb avatar Nov 09 '21 04:11 ramziabb

Hey @ramziabb,

If you like, you can just convert the data.table at the end of your Spectre workflow into a SingleCellExperiment object and go straight into your CATALYST workflow. We were writing up short tutorial for how to do it, but we didn't end up finishing it, but we can send through a quick set of instructions?

tomashhurst avatar Dec 17 '21 06:12 tomashhurst

Hey @tomashhurst would you consider exporting a data.table <--> SingleCellExperiment conversion function in Spectre? Might be quite nice for interchangeable use.

nilseling avatar Dec 17 '21 08:12 nilseling

Hey Nils!! Yes absolutely, we actually have that (and for Seurat etc) in draft format we just haven't had the chance to get it online.

Get Outlook for Androidhttps://aka.ms/AAb9ysg


From: Nils Eling @.> Sent: Friday, December 17, 2021 7:34:47 PM To: HelenaLC/CATALYST @.> Cc: Thomas Ashhurst @.>; Mention @.> Subject: Re: [HelenaLC/CATALYST] Integrating with Spectre Normalize for Batch Correction (Issue #239)

Hey @tomashhursthttps://github.com/tomashhurst would you consider exporting a data.table <--> SingleCellExperiment conversion function in Spectre? Might be quite nice for interchangeable use.

— Reply to this email directly, view it on GitHubhttps://github.com/HelenaLC/CATALYST/issues/239#issuecomment-996532345, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ACZYS6YOQCMZWEXCXVPI6SLURLY2PANCNFSM5HUJY2LQ. You are receiving this because you were mentioned.Message ID: @.***>

tomashhurst avatar Dec 17 '21 08:12 tomashhurst

Awesome! That's really nice!

nilseling avatar Dec 17 '21 08:12 nilseling

Hi guys! Were you able to put together a short tutorial or code fragment to do this? Right now I write out the FCS from Spectre and then back into CATALYST.

Thanks!

ramziabb avatar Feb 09 '22 00:02 ramziabb

The tutorial to convert data.table into SingleCellExperiment object and Seurat object by @tomashhurst is here: https://immunedynamics.io/spectre/tutorials/datatable_interoperability/sce_support.html

Hi, @ramziabb -- I am also trying to analyze CyTOF and flow combing Spectre and Catalyst with batch correction. Would you be happy to share the code for batch correction part?

Have ya looked into https://biosurf.org/cyCombine_detect_batch_effects.html?

My email is [email protected]

Thank you!!!

denvercal1234GitHub avatar Jul 12 '24 15:07 denvercal1234GitHub

Closing as there doesn‘t seem to be an issue per se, but please feel free to continue open discussion, sharing ideas, linking relevant resources 🙏

HelenaLC avatar Jul 15 '24 09:07 HelenaLC