Hypothetical Clustering of a THO Dataset. (a) A small dataset that consistent Trait Heterogeneity Only (THO) genetic model (see Figure 11) is presented. All individuals with the high risk genotype (BB) at locus 1 have Trait I, and all individuals with the high risk genotype (BB) at locus 2 have Trait II. Some individuals have both high risk genotypes and, therefore, both traits. A successful clustering of this dataset might be one in which there are two clusters (b) and (c), such that one cluster contains only individuals who have Trait I (b) and the other cluster contains only individuals who have Trait II (c).
Thornton-Wells et al. BMC Bioinformatics 2006 7:204 doi:10.1186/1471-2105-7-204