Inferring Gene Interactions from Expression Data
Author Information
Author(s): Javier Herrero, Ramón Díaz-Uriarte, Joaquín Dopazo
Primary Institution: Centro Nacional de Investigaciones Oncológicas (CNIO)
Hypothesis
Can we infer transcriptional regulation among genes using large-scale expression data?
Conclusion
The proposed method successfully infers significant gene interactions from time-lagged correlations in expression data.
Supporting Evidence
- The method reduced the dataset dimensionality and extracted significant dynamic correlations.
- 381 significant correlations were found among the gene clusters.
- The network connectivity followed a power law typical of scale-free networks.
- Biological meaning was supported by the distribution of gene ontology terms.
- Connected genes were often transcription factors or had DNA binding domains.
Takeaway
This study shows how scientists can figure out how genes talk to each other by looking at their activity over time.
Methodology
The study used time-course experiments and clustered gene expression profiles to infer gene interactions.
Potential Biases
Potential biases from multiple testing and the assumptions made in the correlation analysis.
Limitations
The method may not distinguish direct from indirect gene interactions.
Participant Demographics
Yeast genes were studied, specifically 6178 genes across complete cell cycles.
Statistical Information
P-Value
< 0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website