Haplotype Association Mapping by Density-Based Clustering in Case-Control Studies (Work-in-Progress)

Linkage disequilibrium (LD) mapping for complex diseases using haplotypes has been intensively studied recently due to increased availability of large-scale dense SNP (single nucleotide polymorphism) markers. Such an LD mapping has many applications, e.g. finding disease-associated haplotypes and predicting disease susceptibility (DS) gene loci from a whole genome scan. In this research, we develop a new algorithmic method for haplotype mapping based on a density-based clustering algorithm, and propose a new haplotype (dis)similarity measure. The mapping regards haplotype segments as data points in a high dimensional space. The DS gene embedded haplotypes, especially those mutants of recent origin, tend to be close to each other due to linkage disequilibrium, while other haplotypes can be regarded as random noise sampled from the haplotype space. Clusters are then identified using a density-based clustering algorithm. Pearson χ statistic or Z-score based on the numbers of cases and controls in a cluster can be used as an indicator of the degree of association between the cluster and the disease under study. The method does not require any assumptions about the evolutionary model or the inheritance patterns of the disease. The proposed similarity measure is a generalization of several haplotype similarity measures currently used in the literature. It is robust against recent mutations/genotype errors and recombination events. Preliminary experimental results on an independent simulated data set, including both SNP markers and microsatellite markers, and on a real data set with the known DS gene location for type 1 diabetes show that our method could predict gene locations with high accuracy, even when the rate of phenocopies is high. This work is still in progress and more data are going to be tested.