A general question, as I’m not sure I can frame the question properly: I have a set of controls (human) and a set of affected patients. In this set of data are also people that are kinda like the controls, kinda like the patients. I thought I was nearly done with the analysis, as I had treated all the samples the same in one big set: imported into R; transformed the data; cleaned it up; determined several ways to get an idea of the number of clusters; ran Harmony, with subsequent UMAP, tSNE, and heatmap plots showing all the data together, as well as the patient data and the “mix” data. Then I’m told that they want to see the data differently: I am to do the controls first, and then map the patient and ‘mix’ data onto the control data, so that the patient and ‘mix’ data take on the same ... characteristics? mapping? ?? ... as the controls.

I’ve tried searching but can’t seem to find anything. I know it is there (it exists) because apparently it is a common thing. The use of R at work hasn’t been applied to cytometry before.

I found the package ‘symphony’ but I keep running out of vector memory. I increased it to 150GB earlier (M1 Mac Studio, 32GB RAM hard-wired, 512GB drive) in Terminal. I have tried with a smaller bit of my data. I’m going to try again with an even more truncated data set. In the meantime, can anyone direct me to a package, or heck, even what words to use in a search? The data is spectral flow cytometry if that makes a difference.

Thanks,