A likelihood approach to classifying fluorescent events collected by multicolor flow cytometry.

Area of Science:

• Microbiology
• Biotechnology
• Analytical Chemistry

Background:

• Flow cytometry is vital for enumerating fluorescently labeled microbes.
• Distinguishing cellular events from noise, especially fluorescent non-cellular particles, is challenging.
• Current methods use arbitrary thresholds, leading to misclassification and unreliable cell proportion estimation.

Purpose of the Study:

• To develop an objective method for separating signal from noise in flow cytometry data.
• To improve the robust classification of fluorescently labeled microbial cells.
• To present a software package for implementing the Z-scoring method.

Main Methods:

• Introduced the Z-scoring approach to objectively separate signal from noise.
• Utilized Gaussian distribution of signal strength to locate noise thresholds for individual fluorophores.
• Employed a likelihood framework to predict fluorescent genotypes and normalize cell counts, enabling robust classification.

Main Results:

• Successfully separated cellular events from noise without arbitrary thresholds.
• Achieved robust and reliable classification of fluorescent cells, including those with multiple fluorophores.
• Demonstrated the ability to classify multiple fluorophores using a single detector.

Conclusions:

• Z-scoring provides an objective and reliable method for flow cytometry analysis of microbial populations.
• The developed method enhances the accuracy of enumerating fluorescently labeled cells in environmental samples.
• The associated software package facilitates the application of Z-scoring for improved flow cytometry data analysis.