A knowledge-guided deep learning framework for quantitative nucleic acid testing

Area of Science:

• Biotechnology
• Molecular Diagnostics
• Infectious Disease Surveillance

Background:

• Nucleic acid testing is crucial for disease screening, genetic analysis, and prenatal diagnosis.
• Current methods rely on optical detection and fluorescence quantitative polymerase chain reaction (PCR) for target amplification.
• Assessing fluorescence values presents challenges due to variations in sample types, reagents, and detection systems.

Purpose of the Study:

• To develop a universal and personalized model for nucleic acid testing result assessment.
• To establish a relationship between fluorescence values and target sequence replication rates.
• To improve the accuracy and generality of nucleic acid detection across different scenarios.

Main Methods:

• Development of a fluorescence value-target sequence replication rate relationship model.
• Utilizing fluorescence quantitative polymerase chain reaction (PCR) for gene amplification.
• Characterizing the relationship between optical signals and biochemical reactions in nucleic acid detection.

Main Results:

• The proposed model accurately assesses nucleic acid testing results based on fluorescence values.
• The model demonstrates both personalized and universal applicability for different testing requirements.
• It addresses challenges in large-scale screening and diverse detection scenarios.

Conclusions:

• A novel fluorescence value-target sequence replication rate model enhances nucleic acid testing accuracy.
• The model provides a versatile solution for personalized and universal assessment in diagnostics.
• This approach is vital for reliable infectious disease detection and control.