Dynamic Bayesian Networks for Gene Expression Data
Author Information
Author(s): Norbert Dojer, Anna Gambin, Andrzej Mizera, Bartek Wilczyński, Jerzy Tiuryn
Primary Institution: Institute of Informatics, Warsaw University
Hypothesis
Can dynamic Bayesian networks be effectively applied to gene expression data from perturbation experiments?
Conclusion
Using data from perturbation experiments significantly improves the quality of inferred gene interaction networks.
Supporting Evidence
- The exact algorithm proposed improves the quality of inferred networks compared to heuristic methods.
- The study demonstrates that perturbation data provides rich information about causal relationships.
- The methodology allows for better handling of gene interactions in time series data.
Takeaway
This study shows that when scientists change how genes work in experiments, they can learn more about how genes interact with each other.
Methodology
The study extends dynamic Bayesian networks to incorporate perturbations in gene expression data and proposes a new discretization method for time series data.
Potential Biases
Potential biases may arise from the simplifications made in the mathematical model used for simulations.
Limitations
The study's findings may not generalize to larger networks due to the small size of the gene set used.
Statistical Information
P-Value
p<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