AutoSpectral improves spectral flow cytometry accuracy through optimised spectral unmixing and autofluorescence-matching at the cellular level

Abstract

The advent of spectral flow cytometry has seen a rapid rise in the complexity of flow cytometry experiments, allowing the construction of assays with at least 50 fluorescent parameters. To correctly determine the contributions of each fluorophores signal to the high parameter data an accurate unmixing matrix needs to be generated. Even with single-stained controls, however, these matrixes include errors such as spillover spread, which compounds with each additional parameter, functionally limiting panel design. An additional source of errors is heterogeneity of cellular autofluorescence, which can affect both the unmixing matrix and misalign signals when the matrix is applied to individual cells in complex cell mixtures. Here we developed AutoSpectral, a statistical approach to automate the production of minimal-residual error unmixing matrixes and pair multiple distinct multifluorescent spectra to individual cells within a mixed sample, via an R-based software tool. AutoSpectral improves unmixing accuracy, improving incorrectly assigned cell positions by up to 9000-fold, reduces spread, particularly in samples with variable autofluorescence, and allows multi-lineage analysis of mixed populations, providing superior data for spectral flow cytometry experiments.