batch correction without ref control workflow -request for link

[klai001]

Hello,
ive saw from the tutorial that the batch correction requires some ref control samples.
however, my samples are randomised across the batches, i dont have a specific sample which i evaluate
across all my batch runs.

here in this link, there's a mention of the workflow for batch integration without ref control. but im unable to find the page with the relevant tutorial.
https://wiki.centenary.org.au/display/SPECTRE/Discovery+workflow+with+batch+alignment+using+CytoNorm

Could you kindly share w me the link for it?
thanks

[josephangelolopez]

Also in the same boat! If there's a workflow for batch integration without a reference control, would be great to give it a try!

[annabellenicolasmith]

I can't speak for the official Spectre team, but I found a way around this by using cyCombine (from biosurf) between the transformation and clustering steps in Spectre. I hope this helps!

[betgou]

Thanks! Just found cyCombine on another thread so its reassuring to hear you've managed it. Thank you!

[betgou]

Do you have an example R script incorporating the cell.dat file @annabellenicolasmith ?

[ghar1821]

Sorry for the radio silence. I recently changed lab and only now have the chance to hop back on Spectre.

We've written up run.cycombine and run.rpca functions, but they are all in the v2-dev branch (https://github.com/ImmuneDynamics/Spectre/tree/v2-dev). The functions should integrate well to existing Spectre based script. Please have a go and let us know if there are issues or bugs. I'll try to find the time to write up a simple tutorial over this weekend or the next one.

You can either install the v2-dev branch using remotes::install_github("ImmuneDynamics/Spectre", ref="v2-dev") or just source the respective files (https://github.com/ImmuneDynamics/Spectre/blob/v2-dev/R/run.cycombine.R) or (https://github.com/ImmuneDynamics/Spectre/blob/v2-dev/R/run.rpca.R) and run the functions.

[betgou]

Great, thank you so much! Will have a look at this tomorrow and report back. Really appreciate your time.