An en masse phenotype and function prediction system for Mus musculus
2008

Predicting Gene Functions in Mice

Sample size: 21603 publication Evidence: high

Author Information

Author(s): Murat Taşan, Weidong Tian, David P. Hill, Francis D. Gibbons, Judith A. Blake, Frederick P. Roth

Primary Institution: Harvard Medical School

Hypothesis

Can a combined approach of guilt-by-profiling and guilt-by-association improve gene function predictions for Mus musculus?

Conclusion

The study achieved high prediction performance for gene functions in mice, with over 80% accuracy for novel predictions.

Supporting Evidence

  • The combined approach outperformed individual methods in predicting gene functions.
  • Over 40% precision was achieved at 1% recall for nearly every GO term.
  • More than 80% of manually examined novel predictions were accurate.

Takeaway

The researchers created a system to guess what genes do in mice by looking at lots of data, and it worked really well!

Methodology

A machine-learning approach combining guilt-by-profiling and guilt-by-association was used to predict gene functions and phenotypes.

Potential Biases

Potential biases may arise from the reliance on existing data and the methods used for predictions.

Limitations

The predictions may not be accurate for all genes, especially those with low existing annotations.

Participant Demographics

The study focused on 21,603 Mus musculus genes.

Digital Object Identifier (DOI)

10.1186/gb-2008-9-s1-s8

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication