Quantitative inference of gene function from diverse large-scale datasets
2008

Automated Gene Function Inference from Large Datasets

publication Evidence: moderate

Author Information

Author(s): Timothy R Hughes, Frederick P Roth

Primary Institution: University of Toronto and Harvard Medical School

Hypothesis

Can gene functions be inferred from large-scale datasets using machine learning methods?

Conclusion

The study demonstrates that gene functions can be effectively predicted using various machine learning approaches applied to large datasets.

Supporting Evidence

  • Automated inference of molecular function of gene products is a key theme in the study.
  • Machine learning methods were used to integrate diverse datasets for gene function prediction.
  • High precision of predictions for many GO terms was achieved using available data sources.

Takeaway

Scientists can guess what genes do by looking at lots of data instead of testing each one in the lab.

Methodology

The study used machine learning methods to integrate thousands of variables describing genes and gene-gene relationships to infer Gene Ontology terms.

Potential Biases

The computational nature of function prediction may introduce biases due to the limited number of evaluated approaches.

Limitations

Participants did not have access to GO annotations from other species, which may limit the understanding of why certain strategies worked well.

Digital Object Identifier (DOI)

10.1186/gb-2008-9-s1-s1

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication