Identifying Blocks and Sub-populations in Noisy SNP Data
We study several problems arising in haplotype block partitioning. Our objective function is the total number of distinct haplotypes in blocks. We show that the problem is NP-hard when there are errors or missing data, and provide approximation algorithms for several of its variants. We also give an algorithm that solves the problem with high probability under a probabilistic model that allows noise and missing data. In addition, we study the multi-population case, where one has to partition the haplotypes into populations and seek a different block partition in each one. We provide a heuristic for that problem and use it to analyze simulated and real data. On simulated data, our blocks resemble the true partition more than the blocks generated by the LD-based algorithm of Gabriel et al. . On single-population real data, we generate a more concise block description than extant approaches, with better average LD within blocks. The algorithm also gives promising results on real 2-population genotype data.
Keywordshaplotype block genotype SNP sub-population stratification algorithm complexity
Unable to display preview. Download preview PDF.
- 2.Bafna, V., Halldorsson, B.V., Schwartz, R., Clark, A., Istrail, S.: Haplotyles and informative SNP selection algorithms: Don’t block out information. In: Proc. of RECOMB, pp. 19–27 (2003)Google Scholar
- 3.Clark, A.: Inference of haplotypes from PCR-amplified samples of diploid populations. Molecular Biology and Evolution 7(2), 111–122 (1990)Google Scholar
- 6.Eskin, E., Halperin, E., Karp, R.M.: Large scale reconstruction of haplotypes from genotype data. In: Proc. of RECOMB, pp. 104–113 (2003)Google Scholar
- 11.Gusfield, D.: Haplotyping by pure parsimony. Technical Report UCDavis CSE- 2003-2, To appear in the Proceedings of the 2003 Combinatorial Pattern Matching Conference (2003)Google Scholar
- 13.Hubbell, E.: Finding a parsimony solution to haplotype phase is NP-hard. Personal’s communicationGoogle Scholar
- 14.Koivisto, M., et al.: AnMDL method for finding haplotype blocks and for estimating the strength of haplotype block boundaries. In: Proc. PSB 2003 (2003)Google Scholar
- 15.MacQueen, J.: Some methods for classification and analysis of multivariate observations. In: Proceedings of the 5th Berkeley Symposium on Mathematical Statistics and Probability, pp. 281–297 (1965)Google Scholar
- 17.Sachidanandam, R., et al.: A map of human genome sequence variation containing 1.42 million single nucleotide polymorphisms. Nature 291, 1298–2302 (2001)Google Scholar