A Compressed Sensing Approach to Pooled RT-PCR Testing for COVID-19 Detection

Area of Science:

• Biotechnology
• Computational Biology
• Epidemiology

Background:

• Traditional pooled testing methods for infectious diseases like COVID-19 can be resource-intensive.
• Quantitative Reverse Transcription Polymerase Chain Reaction (RT-PCR) offers a sensitive detection mechanism.
• Optimizing pooled testing strategies is crucial for efficient large-scale diagnostics.

Purpose of the Study:

• To develop and evaluate 'Tapestry', a single-round pooled testing method for COVID-19 detection.
• To demonstrate Tapestry's ability to reduce testing time and conserve reagents and testing kits.
• To assess Tapestry's performance against established methods like Dorfman pooling.

Main Methods:

• Tapestry integrates principles from compressed sensing and combinatorial group testing.
• It utilizes quantitative RT-PCR readouts for deconvolution of pooled samples.
• Deterministic binary pooling matrices based on Kirkman Triple Systems are proposed for practical implementation.

Main Results:

• Tapestry achieves high accuracy with clinically acceptable false positive/negative rates.
• The method requires significantly fewer tests compared to traditional approaches, especially at low prevalence.
• Empirical evaluations show Tapestry is nearly twice as efficient as two-round Dorfman pooling.

Conclusions:

• Tapestry offers an efficient and resource-saving alternative for large-scale COVID-19 testing.
• The method is viable across a range of prevalence rates, up to 9.5%.
• Tapestry demonstrates strong potential for practical deployment in diagnostic settings.