PSMIX: An R Package for Population Structure Inference
Author Information
Author(s): Wu Baolin, Liu Nianjun, Zhao Hongyu
Primary Institution: University of Minnesota, University of Alabama at Birmingham, Yale University School of Medicine
Hypothesis
Can maximum likelihood methods provide efficient and accurate inference of population stratification and individual admixture compared to Bayesian methods?
Conclusion
PSMIX has similar accuracy to Bayesian methods but offers more efficient computations.
Supporting Evidence
- PSMIX was compared to STRUCTURE and showed similar accuracy.
- PSMIX runs were significantly faster than STRUCTURE runs.
- The correlation coefficient between PSMIX and STRUCTURE results was greater than 0.999 for some datasets.
Takeaway
PSMIX is a computer program that helps scientists understand how different populations are related by looking at their genes, and it does this faster than some other programs.
Methodology
Developed an R package using maximum likelihood methods and the expectation-maximization algorithm.
Limitations
The assumption of independence among markers may affect results.
Participant Demographics
Included individuals from various populations such as Pima, Surui, Sardinian, and Russian.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website