Data quality assessment of ungated flow cytometry data in high throughput experiments

Area of Science:

• Computational Biology
• Bioinformatics
• Data Science in Life Sciences

Background:

• Semiautomated techniques in flow cytometry present data analysis and display challenges.
• Robust quality control and quality assessment are essential for high throughput technologies.
• Bottlenecks exist in processing and analyzing high throughput flow cytometry data, particularly for quality assurance.

Purpose of the Study:

• To introduce graphical exploratory data analytic tools for analyzing ungated flow cytometry data.
• To address the critical need for effective data quality control and assessment in high throughput flow cytometry.

Main Methods:

• Development and implementation of specialized functions within the Bioconductor package 'rflowcyt'.
• Utilizing a variety of graphical exploratory data analytic tools for data exploration.
• Demonstration of methods using two independent sets of high throughput flow cytometry data.

Main Results:

• Graphical representations effectively reveal significant nonbiological variations between samples.
• Empirical Cumulative Distribution Function (ECDF) plots and summary scatterplots are highly effective.
• These visualizations rapidly identify data quality issues missed by manual review.

Conclusions:

• Graphical exploratory data analytic tools provide rapid and effective data quality assessment.
• Proposed visualizations serve as valuable quality assessment tools for high throughput flow cytometry.
• These graphical methods should be integrated into quality control workflows where applicable.