search

RGLab / flowStats

flowStats: algorithms for flow cytometry data analysis using BioConductor tools
15 stars 10 forks source link

Use of warpSet or gaussNorm for CyTOF mass cytometry data #32

Open ashipde opened 4 years ago

ashipde commented 4 years ago

Can one use the warpSet or gaussNorm function of flowStats to normalize high-dimensional CyTOF data? I have such data with a 36-marker panel for 70 samples and I want to reduce batch effects among the samples through normalization.

SamGG commented 4 years ago

I think Greg will bring a solid answer. I just want to point you to the two recent articles on that subject and my post on their approach. http://cytoforum.stanford.edu/viewtopic.php?f=3&t=1875 If you are interested in gaussNorm, Sofie VG has a script that she probably set up during her analysis. I didn't evaluate this as the distribution around the zero looks like a truncated gaussian in the best of the case. I think gaussNorm should be used to realign the positive peak, but the zero stays at zero. https://github.com/saeyslab/CytoNorm_Figures/blob/master/GaussNorm.Rmd Don't hesitate to share your feedback on Cytoforum. Best.

gfinak commented 4 years ago

I agree, I recommend a more modern method. We are not really supporting gaussNorm or warpSet any more.