High Fidelity Machine-Learning-Assisted False Positive Discrimination in Loop-Mediated Isothermal Amplification Using...

Area of Science:

• Biotechnology
• Molecular Biology
• Nanotechnology

Background:

• Loop-mediated isothermal amplification (LAMP) is a valuable nucleic acid detection method for point-of-care applications.
• LAMP assays can be compromised by false positives arising from contamination, nonspecific amplification, or signal reporting.
• Existing methods to improve specificity, such as dye-labeled probes, risk inhibiting the reaction.

Purpose of the Study:

• To develop a probe-free nanopore-based system for LAMP readout.
• To differentiate true LAMP amplicons from false positives using amplicon sizing and counting.
• To enhance the specificity and reliability of LAMP assays.

Main Methods:

• Developed a kinetic model to analyze LAMP reaction patterns.
• Utilized gel electrophoresis to distinguish true from false positive amplicon profiles.
• Implemented nanopore sensing for amplicon sizing and counting, analyzing event charge deficit (ECD) and frequencies.
• Integrated machine learning algorithms for false positive discrimination.

Main Results:

• Distinct kinetic and amplicon size patterns were identified between true and false LAMP positives.
• The nanopore-based sizing and counting method, coupled with machine learning, achieved 91.67% accuracy in discriminating false positives.
• This approach offers a reliable, probe-free readout for LAMP assays.

Conclusions:

• Nanopore-based amplicon sizing and counting provide a robust method for enhancing LAMP specificity.
• This probe-free system overcomes the limitations of dye-based detection, avoiding reaction inhibition.
• The developed system significantly improves the reliability of nucleic acid detection using LAMP technology.