Group Testing Matrix Design for PCR Screening with Real-Valued Measurements.

Area of Science:

• Computational Biology
• Genomics
• Molecular Diagnostics

Background:

• Group testing strategies are crucial for efficient large-scale screening.
• Existing methods like combinatorial group testing and compressed sensing have limitations in matching real-world PCR testing pipelines.
• Efficiently designing measurement matrices is key to accurate positive sample identification.

Purpose of the Study:

• To introduce and validate a real-valued group testing framework for PCR-based diagnostics.
• To develop conditions and algorithms for constructing effective measurement matrices in this new setting.
• To demonstrate the superiority of the proposed method over existing approaches in detecting positive samples.

Main Methods:

• Development of a real-valued group testing model.
• Derivation of conditions for unambiguous decoding of positive samples.
• Algorithm for constructing measurement matrices for small matrix sizes.
• Validation using simulated datasets and wet laboratory experiments with SARS-CoV-2 samples.

Main Results:

• The proposed real-valued group testing framework closely matches PCR testing characteristics.
• Conditions guaranteeing unambiguous decoding were established.
• An algorithm for matrix construction was proposed and evaluated.
• Simulated data showed higher recovery rates of positive samples compared to combinatorial group testing, particularly at higher positivity rates.
• Wet lab experiments validated the approach using real SARS-CoV-2 samples.

Conclusions:

• Real-valued group testing offers a more compatible and effective approach for PCR-based sample pooling.
• The developed matrix design and decoding conditions improve the accuracy and efficiency of disease detection.
• This method shows significant promise for enhancing diagnostic throughput and accuracy in public health surveillance.