An R package implementation of multifactor dimensionality reduction
2011

New R Package for Multifactor Dimensionality Reduction

Sample size: 250 publication Evidence: moderate

Author Information

Author(s): Winham Stacey J, Motsinger-Reif Alison A

Primary Institution: Department of Statistics, North Carolina State University

Hypothesis

The new R package 'MDR' will provide a flexible implementation of the Multifactor Dimensionality Reduction method for variable selection in genetic studies.

Conclusion

The 'MDR' package offers a flexible software solution for R users to identify potential gene-gene interactions.

Supporting Evidence

  • The MDR method has been successful in identifying gene-gene interactions in real data applications.
  • The package is designed to improve the usability of the MDR method for a broader range of users.
  • The package includes options for internal validation and functions for summarizing model fit.

Takeaway

This study introduces a new tool that helps scientists find connections between genes using data from many genetic markers.

Methodology

The package implements the MDR method for variable selection using nonparametric techniques and includes internal validation methods like k-fold cross-validation.

Limitations

The package is primarily suitable for smaller datasets and may not perform well with high-dimensional data due to R's memory limitations.

Participant Demographics

The study used a simulated dataset of 250 individuals genotyped at 25 SNPs.

Digital Object Identifier (DOI)

10.1186/1756-0381-4-24

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication