An Integrative Genomic Approach to Uncover Molecular Mechanisms of Prokaryotic Traits
2006

Understanding Bacterial Traits Through Genomic Analysis

Sample size: 59 publication Evidence: high

Author Information

Author(s): Liu Yang, Li Jianrong, Sam Lee, Goh Chern-Sing, Gerstein Mark, Lussier Yves A

Primary Institution: University of Chicago

Hypothesis

By automatically and simultaneously merging and analyzing massive quantities of microbiological phenotypes and their molecular datasets, we could predict both the molecular underpinnings of prokaryotic phenotypes as well as the relationships between related groups of phenotypes.

Conclusion

This study developed a high throughput computational approach that successfully integrates clinical microbiological data with genomic datasets to uncover the molecular mechanisms underlying bacterial phenotypes.

Supporting Evidence

  • The study identified 3,711 significant correlations between 1,499 distinct Pfam families and 63 phenotypes.
  • Manual evaluation of a random sample showed a minimal precision of 30%.
  • Ten significant correlations between phenotypes and KEGG pathways were unveiled, with eight corroborated in the evaluation.

Takeaway

The researchers figured out how to connect the traits of bacteria to their genes using a computer program, helping us understand how bacteria work better.

Methodology

The study used a high-throughput computational approach to integrate and analyze microbiological phenotypes with genomic datasets across multiple biological scales.

Potential Biases

Potential biases may arise from the overrepresentation of certain bacterial species in the dataset and the limitations of the phenotypic data available.

Limitations

The study's reliance on a specific clinical microbiological database may introduce biases, and the method primarily used sequence-based classifications.

Participant Demographics

The study focused on 59 fully sequenced prokaryotic species, covering various taxonomic groups.

Statistical Information

P-Value

<0.05

Confidence Interval

95% confidence interval: 20%–42%; n = 50

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1371/journal.pcbi.0020159

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication