Predicting Yeast Gene Functions from Network Connections
Author Information
Author(s): Kriston L. McGary, Insuk Lee, Edward M. Marcotte
Primary Institution: University of Texas at Austin
Hypothesis
Loss-of-function phenotypes of yeast genes can be predicted from the loss-of-function phenotypes of their neighbors in functional gene networks.
Conclusion
The study demonstrates that diverse yeast gene loss-of-function phenotypes are predictable from their network connections, which could also be applied to predict human disease genes.
Supporting Evidence
- The study tested 1,102 loss-of-function phenotypes from genome-wide assays of yeast.
- 70% of the phenotypes were predictable at AUC above 0.65.
- The method was validated by predicting and verifying genes whose disruption elongates yeast cells.
Takeaway
Scientists can guess how yeast genes will behave when they are turned off by looking at how other connected genes behave.
Methodology
The study used guilt-by-association in a functional gene network to predict yeast gene essentiality and loss-of-function phenotypes.
Potential Biases
Potential biases may arise from the reliance on existing functional networks and the quality of the data used to construct them.
Limitations
The predictability may be limited by the quality of the seed sets and the inherent noise in the data.
Participant Demographics
The study focused on yeast genes, specifically Saccharomyces cerevisiae.
Statistical Information
P-Value
p<0.01
Confidence Interval
95%
Statistical Significance
p<0.01
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website