4.7.2. Visualization Methods

T-Distributed Stochastic Neighbor Embedding (tSNE) is a nonlinear dimensionality reduction algorithm that reduces high-dimensional data down to two dimensions for easy visualization and rapid exploratory data analysis of any data type [13,52]. To visualize the expression pattern of the panel markers in two dimensions, tSNE maps were plotted in Cytobank (Figure 1b,d). The main cell populations were defined by manual gating according to marker expression and overlaid on the corresponding tSNE plot (Figure 1b).

To further compare the effects of the six dissociation methods on the individual panel markers (and interpatient differences in the individual markers), live singlet cells were gated out according to the Gaussian parameters after CATALYST debarcoding of .fcs files. All live single cells from all files generated from the dissociated patient samples (n = 18) were then grouped separately according to the dissociation method and patient tissue used, and histograms displaying x/5 arcsin-transformed median expression of marker distributions were plotted in MATLAB v.R2019a (The MathWorks Inc., Natick, MA, USA) (Figure S5).