DNEA: an R package for fast and versatile data-driven network analysis of metabolomics data
2024

DNEA: An R Package for Analyzing Metabolomics Data

Sample size: 1818 publication Evidence: high

Author Information

Author(s): Patsalis Christopher, Iyer Gayatri, Brandenburg Marci, Karnovsky Alla, Michailidis George

Primary Institution: University of Michigan

Hypothesis

The DNEA R package improves the analysis of metabolomics data by providing a flexible and powerful tool for constructing biological networks and performing enrichment testing.

Conclusion

The DNEA R package is a more flexible and powerful implementation of the previously published software tool, Filigree, allowing for better analysis of large and complex metabolomics datasets.

Supporting Evidence

  • The DNEA R package is the only publicly available tool designed for constructing biological networks from metabolomics data.
  • The package allows for the analysis of datasets containing exogenous, secondary, and unknown compounds.
  • Eight out of eleven subnetworks identified were found to be statistically significant.
  • The DNEA package includes enhancements over previous tools, allowing for better analysis of metabolomics data.

Takeaway

The DNEA R package helps scientists understand how different metabolites interact in the body, which can help in discovering new ways to treat diseases like diabetes.

Methodology

The DNEA R package uses a multi-step workflow including data pre-processing, model tuning, network estimation, consensus clustering, and enrichment analysis.

Potential Biases

Imbalanced sample sizes across experimental groups can lead to over-representation of certain metabolites in the analysis.

Limitations

The package may require significant computing resources for large datasets, and the performance can be affected by the number of metabolites exceeding the number of samples.

Participant Demographics

The study included 1818 subjects, with 414 testing positive for islet auto-antibodies and 50 developing type 1 diabetes.

Statistical Information

Statistical Significance

FDR < 0.05

Digital Object Identifier (DOI)

10.1186/s12859-024-05994-1

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication