Table 1.
Description of eight genome-wide association mapping models.
| Model | Description | References |
|---|---|---|
| Analysis of variance (ANOVA) | Single locus analysis, the null hypothesis of an ANOVA using a single SNP is that there is no difference between the trait means of any genotype group. | Lewis, 2002 |
| General Linear Model (GLM) | Single locus analysis, the GLM uses principle components as covariates in the model to reduce the false positives that arise due to only population structure. | Price et al., 2006 |
| Mixed Linear Model (MLM) | Single locus analysis, the MLM uses principle components and kinship matrix in the model to reduce the false positives that arise from the family relatedness and population structure. | Yu et al., 2006 |
| Compressed MLM (CMLM) | Single locus analysis, the CMLM clusters the individuals into groups and fits genetic values of groups as random effects in the model that improves statistical power compared to regular MLM methods. | Zhang et al., 2010 |
| Enriched CMLM (ECMLM) | Single locus analysis, the ECMLM calculates kinship using several different algorithms and then chooses the best combination between kinship algorithms and grouping algorithms. | Li et al., 2014 |
| Settlement of MLM Under Progressively Exclusive Relationship (SUPER) | Single locus analysis, the SUPER model uses the associated genetic markers (pseudo Quantitative Trait Nucleotides), instead of all the markers, to derive kinship. Whenever a pseudo QTN is correlated with the testing marker, it is excluded from those used to derive kinship. | Wang et al., 2014 |
| Multiple Loci Mixed Linear Model (MLMM) | Multi-locus analysis, the MLMM incorporates a kinship matrix and selected cofactors, performed better with regard to the false-discovery rate and the QTL detection power than a model incorporating only a kinship matrix or only cofactors. | Segura et al., 2012 |
| Fixed and random model Circulating Probability Unification (FarmCPU) | Multi-locus analysis, this model uses a modified MLM method, Multiple Loci Linear Mixed Model (MLMM), and incorporates multiple markers simultaneously as covariates in a stepwise MLM to partially remove the confounding between testing markers and kinship. To completely eliminate the confounding, MLMM is divided into two parts: Fixed Effect Model (FEM) and a Random Effect Model (REM) and uses them iteratively. FEM contains testing markers, one at a time, and multiple associated markers as covariates to control false positives. To avoid model over-fitting in FEM, the associated markers are estimated in REM by using them to define kinship. The P-values of testing markers and the associated markers are unified at each iteration. | Liu et al., 2016 |