An Empirical Multipoint Method for Identifying
Haplotype Blocks/Tagging SNPs with Application to HapMap Data

Lue Ping Zhao
Fred Hutchinson Cancer Research Center - Seattle

Wednesday, November 16th
3:30pm
Moos 2-530
Minneapolis Campus

Abstract:
The human genome is replete with single nucleotide polymorphisms (SNPs), and SNPs within close proximity are typically in high linkage-disequilibrium (LD) in a block-like structure (Gabriel et al. 2002). As genotyping all possible SNPs for genome-wide association studies is not feasible at this time, it is necessary to reduce the number of SNPs that capture most, if not all, genetic variations. Building upon existing methods for detecting haplotype-blocks (HB) in the genome and for identifying haplotype-tagging SNPs (htSNP), we propose an empirical multipoint approach for both tasks: for HB identification, we introduce a multipoint-statistic to test linkage-equilibrium between HBs, and for htSNPs, we described a multipoint-measure of haplotypic diversity to ensure that htSNPs retain nearly all genetic-diversity information. Using ten ENCODE regions, we have explored the statistical properties of both methods, showing that they are efficient and robust. We also contrast them with other existing methods. Finally, we apply this multipoint-based approach to the genomewide data obtained by the HapMap project (The International HapMap Consortium 2003), and highlight some of interesting results.

A social tea will be held at 3:00 P.M. in A434 Mayo. All are Welcome.
For more details contact 612-624-4655 or see http://www.biostat.umn.edu/seminar_academic.html