From correlation to causation networks: a simple approximate learning algorithm and its application to high-dimensional plant gene expression data
2007

Learning Causal Networks from Gene Expression Data

Sample size: 800 publication Evidence: moderate

Author Information

Author(s): Rainer Opgen-Rhein, Korbinian Strimmer

Primary Institution: Ludwig-Maximilians-Universität München

Hypothesis

Can a simple heuristic algorithm effectively infer causal networks from high-dimensional gene expression data?

Conclusion

The proposed heuristic algorithm can efficiently identify causal structures in high-dimensional genomic data, even with small sample sizes.

Supporting Evidence

  • The algorithm yields sensible approximations of causal structures in genomic data.
  • It is computationally efficient and applicable to high-dimensional data.
  • The method is implemented in the 'GeneNet' R package.

Takeaway

This study shows a way to figure out how genes affect each other using a smart method that works even when there aren't many examples to learn from.

Methodology

The method transforms a correlation network into a partial correlation graph and establishes a partial ordering of nodes through multiple testing.

Limitations

The algorithm is based on approximations and may not capture all causal relationships accurately.

Digital Object Identifier (DOI)

10.1186/1752-0509-1-37

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication