Abstract: Objective To investigate the feasibility and limitations of theoretical spectra applied as substitutes for single-stained controls in spectral flow cytometry, and to propose a reference-spectrum optimization strategy for correcting deviations between theoretical and experimentally measured spectra caused by auto-fluorescence. Methods Mouse bone marrow single-stained samples were collected and analyzed by multi-platform spectral flow cytometry. Cosine similarity was used to evaluate the consistency between theoretical spectra and experimentally measured spectra. For single-stained samples affected by autofluorescence, reference spectra were generated from cell subsets with relatively homogeneous auto-fluorescence, and their effects on performance were assessed. Results Among eleven antibody-fluorochromes/fluorescent dyes, five of them showed cosine similarity values of 0.90 or higher between measured and theoretical spectra on both platforms, which suggested that they might serve as alternatives to single-stained controls. Spectral mismatches manifested either as spectral shifts related to detection conditions or, in a subset of samples, as abnormal reference spectra associated with auto-fluorescence interference. The subset strategy corrected abnormal reference spectra caused by auto-fluorescence heterogeneity and significantly reduced unmixing spillover errors in mouse bone marrow cells. Conclusions Theoretical spectra can serve as effective tools for evaluating and correcting reference spectra and may partially replace single-stained controls. The subset strategy can reduce reference-spectrum bias caused by autofluorescence and improve the accuracy of spectral unmixing.

Key words: spectral flow cytometry, theoretical spectra, reference spectra, autofluorescence, subset strategy