Predicting Human Height Using Genetic Information
Author Information
Author(s): Robert Makowsky, Nicholas M. Pajewski, Yann C. Klimentidis, Ana I. Vazquez, Christine W. Duarte, David B. Allison, Gustavo de los Campos
Primary Institution: Department of Biostatistics, University of Alabama at Birmingham
Hypothesis
How can genomic prediction improve the understanding of human height variation?
Conclusion
The study shows that while whole-genome prediction models can explain a significant amount of height variation, there is still considerable room for improvement in predictive accuracy.
Supporting Evidence
- Models using a large number of SNPs can explain a significant proportion of height variation.
- Predictive accuracy increases with the number of close relatives included in the training dataset.
- Whole-genome prediction methods outperform traditional models based on a few selected SNPs.
Takeaway
Scientists are trying to figure out how our genes affect how tall we are, and they found that using a lot of genetic information helps, but there's still more to learn.
Methodology
The study used data from the Framingham Heart Study and applied various statistical models to predict height based on genetic markers.
Potential Biases
Potential bias may arise from overfitting and the use of related individuals in the training dataset.
Limitations
The predictive ability of the models may be limited by the genetic architecture and the relationship between individuals in the training and validation samples.
Participant Demographics
Participants were adults aged 18 and older, with 2,311 males and 2,806 females.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website