Daily Arxiv

This is a page that curates AI-related papers published worldwide.
All content here is summarized using Google Gemini and operated on a non-profit basis.
Copyright for each paper belongs to the authors and their institutions; please make sure to credit the source when sharing.

A U-Statistic-based random forest approach for genetic interaction study

Created by
  • Haebom

Author

Ming Li, Ruo-Sin Peng, Changshuai Wei, Qing Lu

Outline

This paper proposes the Forest U-Test, a novel method for studying the influence of multiple genetic variants, environmental risk factors, and their interactions on the variation in complex traits. To overcome the limitations of existing methods, which involve exponentially increasing feature spaces and computational intensity, we employ a random forest approach based on the U-statistic. Simulations demonstrate that our method outperforms existing methods and, when applied to a study of cannabis dependence (CD), detects significant joint associations (p-value < 0.001) in three independent datasets. These results are also replicated in two independent datasets (p-values of 5.93e-19 and 4.70e-17, respectively).

Takeaways, Limitations

•
Takeaways:
◦
We present a novel method (Forest U-Test) for effectively detecting gene-gene and gene-environment interactions in high-dimensional genetic association studies.
◦
Demonstrated superior performance over existing methods through simulation experiments and real data analysis.
◦
It suggests applicability to the study of complex traits such as drug addiction.
•
Limitations:
◦
Further research may be needed to determine the general applicability of the currently presented method.
◦
Additional validation may be required for various types of complex traits and datasets.
◦
Further research may be needed on parameter optimization of U-statistic-based random forests.
👍
Made with Slashpage