To Friday 07 September 2012
This course provided a general background for statistical models used in association studies with an emphasis on mixed model approaches. Multi-trait and multi-locus extensions of models of genotype-phenotype association were discussed with applications to data and hands-on examples. The course covered both genotype-phenotype associations and genotype-environment associations, and discussed parallels between these approaches. In addition, downstream analyses (e.g., gene set and functional class enrichment and experimental work) has been discussed that can be used to help validate association study results and to clarify the biological basis for the observed signals. Participants received practical experience running analyses with a sample data set using R packages.
1. Introduction to mixed models with application to genotype-phenotype association studies
2. Another application of mixed models: genotype-environment association studies.
3. More complex models such as multi-trait and multi-locus models.
4. Range-wide versus population-specific analyses - pros and cons of using geographically restricted versus range-wide sampling
5. Methods for validating the results and integrating results with other genome-wide information (e.g, enrichment of likely functional variants, relevant candidate gene sets, experimental follow-up) and making biological interpretations
6. Practical experience using R with a sample data set