Related Experiment Video
Updated: Aug 6, 2026
07:11
Digital PCR-based Competitive Index for High-throughput Analysis of Fitness in Salmonella
Published on: May 13, 2019
A Bayesian contamination model for serial dilution assays
Siquan Wang1, Qixuan Chen1, Matthew S Perzanowski2
1Department of Biostatistics, Columbia University, New York, NY 10032, United States.
Biometrics
|July 24, 2026
Summary
This study introduces a Bayesian contamination model to accurately estimate sample concentrations in serial dilution assays. The new method effectively flags contaminated samples, improving concentration estimation accuracy.
Area of Science:
- Biomedical research
- Analytical chemistry
- Statistical modeling
Background:
- Serial dilution assays are crucial in biomedical research.
- Current methods struggle with measurements near calibration curve bounds and are susceptible to sample contamination.
- Contamination can lead to inaccurate concentration estimations.
Purpose of the Study:
- To develop a Bayesian contamination model for serial dilution assays.
- To simultaneously flag contaminated samples and estimate unknown concentrations.
- To quantify estimation uncertainty and improve accuracy.
Main Methods:
- Proposed a Bayesian contamination model.
- Conducted extensive simulation studies under various contamination scenarios.
- Applied the model to raw immunoassay data from a real-world study.
Main Results:
- The proposed Bayesian model consistently outperformed commonly used approaches in simulations.
- The method successfully flagged contaminated samples and provided accurate concentration estimations.
- Demonstrated a practical Bayesian workflow for contamination modeling.
Conclusions:
- The Bayesian contamination model offers a practical and robust solution for detecting contaminated samples in serial dilution assays.
- This approach enhances the accuracy of concentration estimation in immunoassays.
- Presents a generalizable Bayesian framework for handling unknown contamination processes.
