Related Experiment Video
Updated: Jul 3, 2026
09:37
Navigating MARRVEL, a Web-Based Tool that Integrates Human Genomics and Model Organism Genetics Information
Published on: August 15, 2019
An en masse phenotype and function prediction system for Mus musculus
Murat Taşan1, Weidong Tian, David P Hill
1Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Longwood Avenue, Boston, Massachusetts 02115, USA.
Genome Biology
|July 22, 2008
Summary
Researchers can now predict gene function more accurately in mice using an integrated data approach. This method combines guilt-by-profiling and guilt-by-association to improve hypothesis generation for biological research.
Area of Science:
- Genomics
- Bioinformatics
- Computational Biology
Background:
- High-throughput biological data presents challenges for individual researchers.
- Integrating diverse data sources is crucial for hypothesis generation.
Purpose of the Study:
- To predict Gene Ontology (GO) terms and phenotypes for Mus musculus genes.
- To evaluate a combined guilt-by-profiling and guilt-by-association method.
Main Methods:
- Applied a previously established prediction method to mouse genes.
- Integrated diverse data including expression, interaction, and sequence data.
- Evaluated predictions using held-out genes and manual literature review.
Main Results:
- Achieved high prediction performance for GO terms, with >40% precision at 1% recall.
- Manually validated over 80% of novel GO term predictions and >40% of phenotype predictions.
- Demonstrated superior performance of the combined approach over individual methods.
Conclusions:
- The integrated approach provides accurate gene function and phenotype predictions for Mus musculus.
- This method enhances the ability to form accurate hypotheses from complex biological data.
- The combined guilt-by-profiling and guilt-by-association strategy is effective for mouse gene annotation.
