Related Experiment Video
Updated: Dec 14, 2025
10:36
Rare Event Detection Using Error-corrected DNA and RNA Sequencing
Published on: August 3, 2018
12.4K
Bayesian modelling of high-throughput sequencing assays with malacoda
Andrew R Ghazi1, Xianguo Kong2, Ed S Chen3
1Quantitative and Computational Biosciences, Baylor College of Medicine, Houston, Texas, United States of America.
Plos Computational Biology
|July 22, 2020
Summary
A new R package, malacoda, provides a robust Bayesian framework for analyzing next-generation sequencing (NGS) functional genomics screens. This tool addresses statistical challenges in massively parallel reporter assays (MPRAs) to accurately characterize genetic variant function.
Area of Science:
- Genomics
- Bioinformatics
- Statistical Genetics
Background:
- Next-generation sequencing (NGS) has expanded the catalog of human genetic variation, but experimental characterization of variant function remains a challenge.
- High-throughput screens accelerate variant functionalization, yet comprehensive statistical methods for analyzing screen data have lagged.
- Massively parallel reporter assays (MPRAs) measure transcriptional changes from thousands of genetic variants but present statistical complexities like overdispersion and depth dependence.
Purpose of the Study:
- To develop and present an extensive statistical framework for analyzing NGS functionalization screens.
- To introduce the R package 'malacoda' that implements a probabilistic, fully Bayesian model for MPRA data analysis.
- To enable the integration of external annotations for more informative prior estimations and improved data interpretation.
Main Methods:
- Developed a probabilistic, fully Bayesian model using the negative binomial distribution with gamma priors to analyze sequencing counts.
- The model accounts for experimental factors such as input library preparation and sequencing depth.
- Incorporated empirical prior estimation leveraging high-throughput data and allowed integration of external annotations (e.g., ENCODE, DeepSea).
Main Results:
- The 'malacoda' R package offers automated barcode counting, quality control, and visualization functions.
- Validation using literature data, simulated assays, and primary MPRA data demonstrated the method's robustness.
- Luciferase assays confirmed experimental validation of selected variants, including those with method disagreements and those uniquely identified using external annotations.
Conclusions:
- The 'malacoda' package provides a statistically rigorous and comprehensive framework for analyzing NGS functionalization screens, particularly MPRAs.
- This approach addresses key statistical challenges, improves the accuracy of variant function characterization, and facilitates data integration.
- The developed software enhances the utility of high-throughput screening for understanding genetic variation and its functional impact.
