E-Flow

E-Flow processes fluorescence flow cytometry (FCM) data to quantify single-cell gene expression distributions in bacteria.

Key Features:

• Debris Filtering: Uses a Bayesian mixture model that leverages all forward- and side-scattering signals to distinguish debris from viable bacterial cells.
• Autofluorescence and Shot Noise Correction: Implements calibration-based corrections to remove autofluorescence and shot noise biases from fluorescence measurements.
• Scattering Signal Scaling Correction: Provides methods to account for non-linear scaling of forward- and side-scatter with cell size and shot noise to estimate true biological expression means and variances without relying solely on scattering-based size estimates.
• Implementation: Implemented as an R package developed by the van Nimwegen Lab that processes raw FCM data for bacterial single-cell analysis.

Scientific Applications:

• High-throughput single-cell gene expression quantification: Enables precise estimation of means and variances of gene expression distributions in bacterial populations from FCM data.
• Comparative quantitative inference with microscopy: Facilitates systematic comparison of FCM measurements to microscopic setups for rigorous quantitative inference of gene expression dynamics.

Methodology:

Processing of raw FCM data includes removal of non-viable cells via a Bayesian mixture model using scattering signals, calibration-based correction for autofluorescence and shot noise, and methods to correct non-linear scaling of forward- and side-scatter to estimate biological expression means and variances while reducing electronic noise from the cytometer.