Identifying Metabolic Enzymes Using Association Evidence
Author Information
Author(s): Peter Kharchenko, Lifeng Chen, Yoav Freund, Dennis Vitkup, George M Church
Primary Institution: Harvard Medical School
Hypothesis
Can a combination of genome context and functional association evidence effectively predict genes encoding metabolic enzymes?
Conclusion
The study demonstrates that combining genome context and functional association evidence significantly improves the prediction of genes encoding metabolic enzymes.
Supporting Evidence
- The method predicted 60% of enzyme-encoding genes of E. coli within the top 10 candidates.
- Combining multiple types of association evidence led to significantly better predictions than using any single type.
- The approach does not rely on direct sequence homology to known enzyme-encoding genes.
Takeaway
The researchers found a way to guess which genes make enzymes by looking at how genes are related to each other, rather than just comparing their DNA.
Methodology
The study used a novel method that combines various types of functional association evidence, including gene clustering, phylogenetic profiles, and gene expression data.
Potential Biases
Potential bias from excluding paralogous enzyme pairs and reliance on specific datasets.
Limitations
The method may not effectively predict additional functions for already known enzyme-encoding genes.
Participant Demographics
The study focused on metabolic networks of E. coli and S. cerevisiae.
Statistical Information
P-Value
null
Confidence Interval
null
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website