Haplotype Block Partitioning and TagSNP Selection on Human Chromosome 21

Abstract

A Single Nucleotide Polymorphism or SNP is a DNA se- quence variation occurring when a single nucleotide in the genome di®ers between members of species. Recent research reveals that SNPs within certain haplotype blocks induce only a few distinct common haplotypes in the majority of the pop- ulation. The existence of haplotype block structures has se- rious implications for association-based methods in mapping of disease genes. In this paper, we propose and implement several e±cient algorithms for identifying haplotype blocks in the genome. For dealing with missing data, which often appears in real biomedical data, we develop methods that as- sign missing data to speci¯c alleles such that the resulting diversity is minimized. Our developed system is used for analyzing chromosome 21 haplotype data provided by Patil et al. [11]. In contrast to previous partition methods, the haplotype blocks and tagSNPs identi¯ed by our methods are longer and fewer, yet still retaining the desired expressing capability.