Skip to contents

Projects the autofluorescence spectral variants into fluorophore unmixed space to determine which best fits each cell (event). A fast approximation for brute force sequential unmixing method in early versions of AutoSpectral. Provides essentially identical results to minimization of fluorophore signal (dist0 method). Substantially faster. Use L1 (absolute value) minimization, which works better than the standard L2 (squared error).

Usage

assign.af.fluorophores(raw.data, spectra, af.spectra)

Arguments

raw.data

Expression data from raw FCS files. Cells in rows and detectors in columns. Columns should be fluorescent data only and must match the columns in spectra.

spectra

Spectral signatures of fluorophores, normalized between 0 and 1, with fluorophores in rows and detectors in columns.

af.spectra

Spectral signatures of autofluorescences, normalized between 0 and 1, with fluorophores in rows and detectors in columns.

Value

Row indices for best-fitting AF spectra (from af.spectra)