Two-Stage Two-Locus Models in Genome-Wide Association
Author Information
Author(s): David M. Evans, Jonathan Marchini, Andrew P. Morris, Lon R. Cardon
Primary Institution: Wellcome Trust Centre for Human Genetics, University of Oxford
Hypothesis
Can we detect interacting loci in genome-wide association studies using realistic sample sizes?
Conclusion
An exhaustive search involving all pairwise combinations of markers across the genome is more effective than single-locus scans for identifying interacting loci.
Supporting Evidence
- An exhaustive two-locus search is more powerful than a single-locus strategy when loci interact.
- Two-stage strategies can reduce computational burden but may miss important interactions.
- 1,500 individuals are sufficient to detect loci responsible for moderate proportions of trait variance.
Takeaway
The study shows that looking at pairs of genes together can help find important interactions that single-gene tests might miss.
Methodology
The study used simulations to compare the power of single-locus and two-locus models in detecting genetic interactions.
Potential Biases
Potential bias due to the reliance on specific models and assumptions about genetic variance.
Limitations
The study focused on a limited range of models and may not capture all possible genetic interactions.
Participant Demographics
1,500 unrelated individuals were simulated.
Statistical Information
P-Value
p<0.05
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website