I was checking the code for batch effect correction, and I found that “DMSO_P01579”, “DMSO_Q16552”, “DMSO_P05113” were deselected from the data when being normalized. I am aware that the value for those three stimulation+protein_id is either being 1.0 or missing. What is special about them?
Thanks for your question. In T cell polarization assays, DMSO is typically used as a control to establish the baseline state of T cells without any stimulation. In these assays, T cells are subjected to various stimuli to drive their differentiation into subsets like Th1, Th2, or Th17 cells. DMSO acts as a negative control to confirm that any observed changes result from specific stimuli rather than other factors. Since batch effect correction and downstream data integration rely on biological variability across samples, DMSO controls, which show minimal or no variation, are not informative for these processes and were, therefore, removed during data processing.
I hope this answers your query. Feel free to write here if you need more information.
Following on from this question, you also supplied a processed PT_P01579(IFNγ)/PT_P05113(IL5).tsv file. To calculate this ratio, did you use data from the DMSO samples? Also, could you clarify if Q16552 (IL-17A) was used in the calculations?
Thanks for your query. We did not use values from DMSO or the Q16552 (IL-17A) readouts to calculate the PT_P01579 (IFNγ)/PT_P05113 (IL5) ratio. If you would like to use DMSO and Q16552 (IL-17A) readings, please refer to the raw data from the T-cell assay datasets available here.
I hope this answers your question, and please feel free to reach out if you need more information.