Related Experiment Video
Updated: Aug 10, 2025
09:29
Using the Open-Source MALDI TOF-MS IDBac Pipeline for Analysis of Microbial Protein and Specialized Metabolite Data
Published on: May 15, 2019
19.6K
BZINB model-based pathway analysis and module identification facilitates integration of microbiome and metabolome
Bridget Lin1, Hunyong Cho1, Chuwen Liu1
1Department of Biostatistics, Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, United States.
Biorxiv : the Preprint Server for Biology
|February 13, 2023
Summary
A new bivariate zero-inflated negative binomial (BZINB) model improves microbiome-metabolome data analysis by accurately capturing correlations, outperforming traditional methods for understanding health and disease. This approach is crucial for zero-inflated biological data.
Area of Science:
- Computational Biology
- Bioinformatics
- Microbiome Research
Background:
- Integrating multi-omics data (microbiome, metabolome) is vital for understanding human health and disease.
- Existing correlation methods struggle with the excess zeros common in microbiome data.
Approach:
- Introduced a bivariate zero-inflated negative binomial (BZINB) model-based network and module analysis.
- Developed BZINB-iMMPath for constructing metabolite-species and species-species correlation networks.
- Validated the method using simulated and real multi-omics data from a childhood oral health study (ZOE 2.0).
Key Points:
- BZINB model-based correlation analysis demonstrated superior accuracy over Spearman and Pearson correlations for zero-inflated data.
- BZINB-iMMPath effectively identifies modules of correlated species and facilitates group comparisons.
- Analysis of ZOE 2.0 data revealed significant differences in microbial-metabolite correlations between healthy and caries-affected children.
Conclusions:
- The BZINB model provides a robust alternative for analyzing zero-inflated bivariate count data in multi-omics studies.
- This method enhances the integration of microbiome and metabolome data for biological discovery.
- BZINB-iMMPath is suitable for investigating complex biological relationships in health and disease.
