Today’s journal article of the day *fanfare*:
“How To Perform Meaningful Estimates of Genetic Effects.” Alvarez-Castro JM et al. 2008.
The authors discuss the problem inherent in working with models of gene interaction on a genome wide scale and the role of gene interactions in determining genetic traits(epistasis). This article looked at the application of a recently developed model called the NOIA (Natural and Orthogonical InterActions) to provide functional estimates of genetic effects. To accomplish this, the authors applied this new model to an experimental data-set. The results from this model was then compared to other models frequently used that did not account for divergences from the HWE in the same ways NOIA did. The results found suggested that NOIA does better than the other two tested models(G2A and F2) in additive variance of the trait tested in the simulated population.
The method section was way over my head with the math involved to create and explain the model, but the idea seems sound enough. Questions I would have for the authors would be: 1)Why not test the NOIA model against more than just two models? There are many more models out there to use(ex. Bayesian Models), so why stop at two if you are really serious about rigourously testing the new model. 2)Would there be instances where the NOIA model would not be the correct model to use(populations that do adhere to HWE, very small populations, populations that are genetically similar, etc.)?
That’s it for today! Check back again soon for more awesome and exciting articles from the dark realms of scientific literature. 😛