Predicting Protein Function with Hierarchical Phylogenetic Profiles
Author Information
Author(s): Ranea Juan A., Yeats Corin, Grant Alastair, Orengo Christine A.
Primary Institution: University College London
Hypothesis
Can hierarchical phylogenetic profiles improve the prediction of protein functions in eukaryotic genomes?
Conclusion
The study demonstrates that the new phylogenetic profiling method can identify functional relationships in eukaryotic proteins that traditional methods cannot.
Supporting Evidence
- Eukaryotic protein clusters show higher Ro information content than prokaryotic profiles.
- The Phylo-Tuner method can detect functional relationships that traditional methods miss.
- Profiles with Ro ≥ 0.1 were selected for further analysis to focus on multigene families.
- Z scores indicate high precision in distinguishing true positive predictions from false positives.
- Profiles based on domain occurrence can reduce noise from domain rearrangements.
Takeaway
Scientists created a new way to look at proteins in eukaryotes to figure out what they do, which is better than older methods.
Methodology
The study used CATH structural domain assignments from the Gene3D database to analyze 13 complete eukaryotic genomes and developed a novel phylogenetic profiling method based on domain copy number.
Potential Biases
Potential bias in functional predictions due to the correlation of gene copy number with genome size.
Limitations
The method may not accurately identify specific orthologous genes involved in functional associations due to the presence of multiple gene copies.
Participant Demographics
The study analyzed 13 eukaryotic genomes, including various species such as fungi, fish, and mammals.
Statistical Information
P-Value
0.1
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website