Related Experiment Video
Updated: Aug 26, 2025
08:23
Single Droplet Digital Polymerase Chain Reaction for Comprehensive and Simultaneous Detection of Mutations in Hotspot Regions
Published on: September 25, 2018
13.3K
Adaptive Filtering Framework to Remove Nonspecific and Low-Efficiency Reactions in Multiplex Digital PCR Based on
Luca Miglietta1,2, Ke Xu1,2, Priya Chhaya2
1Department of Infectious Disease, Faculty of Medicine, Imperial College London, LondonW12 0NN, U.K.
Analytical Chemistry
|October 3, 2022
Summary
Filtering nonspecific reactions in real-time digital polymerase chain reaction (qdPCR) using outlier detection improves machine learning-based classification. This novel framework enhances data-driven multiplexing for accurate molecular diagnostics.
Area of Science:
- Molecular Diagnostics
- Bioinformatics
- Real-time PCR
Background:
- Real-time digital polymerase chain reaction (qdPCR) combined with machine learning (ML) shows promise for molecular diagnostics.
- Data-driven multiplexing extracts kinetic and thermodynamic information from amplification curves for target classification.
- Undesired amplification events and suboptimal reaction conditions hinder accurate target classification in qdPCR.
Purpose of the Study:
- To develop a novel framework for identifying and filtering nonspecific and low-efficient qdPCR reactions using outlier detection.
- To improve the classification performance of data-driven multiplexing methods, specifically Amplification Curve Analysis (ACA).
- To introduce an Adaptive Mapping Filter (AMF) to dynamically adjust outlier removal based on positive qdPCR counts.
Main Methods:
- Proposed a framework using outlier detection algorithms based on sigmoidal trends of amplification curves to filter qdPCR data.
- Implemented the framework to enhance the ACA method using published data for screening carbapenemase-producing organisms.
- Developed the AMF strategy to adjust the percentage of outliers removed based on the number of positive qdPCR counts.
Main Results:
- The filtering framework identified abnormal amplification curves (outliers) linked to shifted melting distribution or decreased PCR efficiency.
- Application of AMF to ACA improved sensitivity by 1.2% when using inliers versus a 19.6% decrease when using outliers (p < 0.0001).
- The method successfully removed 53.5% of incorrect melting curves based solely on amplification curve shape.
Conclusions:
- Filtering nonspecific and low-efficient reactions significantly enhances classification accuracy in advanced multiplexing techniques.
- Demonstrated a strong correlation between amplification curve kinetics and melting curve thermodynamics.
- The proposed framework and AMF offer a robust approach to improve the reliability of qdPCR-based molecular diagnostics.
