ParaHaplo 3.0: A program package for imputation and a haplotype-based whole-genome association study using hybrid parallel computing
© Misawa and Kamatani; licensee BioMed Central Ltd. 2011
Received: 28 February 2011
Accepted: 24 May 2011
Published: 24 May 2011
Use of missing genotype imputations and haplotype reconstructions are valuable in genome-wide association studies (GWASs). By modeling the patterns of linkage disequilibrium in a reference panel, genotypes not directly measured in the study samples can be imputed and used for GWASs. Since millions of single nucleotide polymorphisms need to be imputed in a GWAS, faster methods for genotype imputation and haplotype reconstruction are required.
We developed a program package for parallel computation of genotype imputation and haplotype reconstruction. Our program package, ParaHaplo 3.0, is intended for use in workstation clusters using the Intel Message Passing Interface. We compared the performance of ParaHaplo 3.0 on the Japanese in Tokyo, Japan and Han Chinese in Beijing, and Chinese in the HapMap dataset. A parallel version of ParaHaplo 3.0 can conduct genotype imputation 20 times faster than a non-parallel version of ParaHaplo.
ParaHaplo 3.0 is an invaluable tool for conducting haplotype-based GWASs. The need for faster genotype imputation and haplotype reconstruction using parallel computing will become increasingly important as the data sizes of such projects continue to increase. ParaHaplo executable binaries and program sources are available at http://en.sourceforge.jp/projects/parallelgwas/releases/.
KeywordsParaHaplo haplotype reconstruction genotype imputation parallel computing HapMap GWAS
Recent advances in various high-throughput genotyping technologies have allowed us to test allelic frequency differences between case and control populations on a genome-wide scale . Genome-wide association studies (GWASs) are used to compare the frequency of alleles or genotypes of a particular variant between cases and controls for a particular disease across a given genome [2–4]. More than a million single nucleotide polymorphisms (SNPs) are analyzed in SNP-based GWASs and haplotype-based GWASs [5, 6].
By modeling the patterns of linkage disequilibrium in a reference panel, genotypes not directly measured in the study samples can be imputed . SNP genotype imputation has been proposed as a powerful means to include genetic markers into large-scale disease association studies without the need to actually genotype them [8, 9]
To quickly conduct GWASs, we developed a software package for the parallel computation of genotype imputation and haplotype reconstruction called ParaHaplo 3.0. ParaHaplo 3.0 contains all of the functions of ParaHaplo 1.0  and ParaHaplo 2.0 , plus it can conduct genotype imputation and haplotype reconstruction using MACH 1.0 . ParaHaplo 3.0 is based on the principle of data parallelism, a programming technique used to split large datasets into smaller ones that can be run in a parallel concurrent fashion . ParaHaplo 3.0 is intended for use in workstation clusters using the Intel Message Passing Interface (MPI).
Using ParaHaplo 3.0, we estimated haplotypes using the genotype data of the Japanese from Tokyo (JPT) and the Han Chinese from Beijing (CHB) obtained from the HapMap dataset [12, 13]. Using ParaHaplo 3.0, we compared the speed of haplotype estimation using parallel computation to the number of processors.
ParaHaplo supports the genotype data in the HapMap format  and the BioBank Japan format . ParaHaplo 3.0 requires an input file of haplotype block boundaries. ParaHaplo 3.0 can conduct genotype imputation and haplotype reconstruction using MACH 1.0 . ParaHaplo 3.0 can also conduct haplotype estimation using PHASE 2.1  and SNPHAP 1.3.1  algorithms. By using hybrid MPI + OpenMP parallelization , ParaHaplo 3.0 can conduct haplotype-based GWAS faster than previous versions.
Parallel computing using MPI methods
ParaHaplo 3.0 is implemented in an MPI-C multithreaded package. The MPI package allows us to construct parallel computing programs on multiprocessors. The genome-wide polymorphism data is broken down into user-defined haplotype blocks, and the MPI Bcast function is used to distribute a single block of haplotype data into each processor. Each processor executes Mach 1.0  and conducts genotype imputation and haplotype reconstruction of a single linkage disequilibrium (LD) block. Once the haplotypes of each LD block are completely estimated, the results are compiled into a single genome-wide dataset through use of the MPI-Gatherv function. ParaHaplo 3.0 is compatible with OpenMPI 1.2.5 and MPICH 1.2.7p1. Users can compile the source code using a GCC compiler, an Intel C compiler, or a Fujitsu C compiler, so that Haplotype-based GWAS can be run on Linux-based PC clusters as well as on K computer (http://www.fujitsu.com/global/news/pr/archives/month/2009/20090717-01.html).
A PC cluster at RIKEN Integrated Cluster of Clusters (RICC) was used when the computational time was measured. The program was compiled using an Intel C compiler. The numbers of processing units used included 1, 2, 4, 8, 16, 32, and 64.
An example GWAS is presented here. We used ParaHaplo 3.0 to compare genome-wide genotype data of JPT and CHB from HapMap 3.0 . Some individuals were excluded because they contain too many untyped SNPs. As the reference panel, we used 20,086 SNPs of 170 people. JPT data set was used to be imputed. JPT data set consisted of 82 people with 2,392 SNPs being untyped. Haplotype blocks were obtained as LD blocks using the method outlined by Gabriel et al.  and the Haploview program . The entire JPT and CHB genomes were divided into 106,149 haplotype blocks by Haploview . Among them, 1,536 haplotype blocks were on chromosome 22.
Results and discussion
Genotype Imputation and Haplotype Reconstruction of JPT and CHB
Elapsed times and speedups obtained using ParaHaplo 3.0 in the imputation process
Number of Processing Units
Parallel Computation of Haplotype-Based GWAS
The results show that the parallel computing ability of ParaHaplo 3.0 for haplotype estimation was 20 times faster than that of the non-parallel version of ParaHaplo 3.0. In this study, we used a total of 89 JPT and CHB individuals whose genotypes had been determined during the HapMap project . When a single processor was used, haplotype estimation for chromosome 22 took more than 4 h; if 9,000 individuals were to be analyzed under the same conditions, the analysis would take more than 2 weeks. However, if ParaHaplo 3.0 was used on a workstation with 64 processors, the same analysis would take less than 1 day. However, the relatively lower speed ratio could be affected by the inflation of sample size . Further study is required.
Even when 64 processors were used, the speedup ratio was only 22 because of the variations in the LD block size. ParaHaplo is based on data parallelism, and our result showed that the computation time of each genotype imputation was approximately proportional to the number of SNPs within the LD block (data not shown); therefore, we believe that a large LD block may create a computational bottleneck as does in haplotype estimation .
We developed ParaHaplo 3.0, a set of computer programs, for the parallel computation of haplotype estimation and accurate P values in haplotype-based GWASs. ParaHaplo is intended for use in workstation clusters using the Intel MPI. Using ParaHaplo, we conducted haplotype estimation of JPT and CHB genotype data taken from the HapMap 3.0 dataset .
These results indicate that when the number of processors is sufficient, the parallel computing abilities of ParaHaplo are 20 times faster than those of non-parallel programs. Accurate and complete genotypes have been obtained for more than a million SNPs , and >10,000 individuals are now being genotyped . The need for fast haplotype estimation using parallel computing will become increasingly important as project data sizes continue to increase.
Availability and Requirements
Project name: ParaHaplo 3.0
Project home page: http://sourceforge.jp/projects/parallelgwas/releases/46982
Operating systems: Platform independent
Programming language: Java and C
Other requirements: OpenMPI version 1.2.5, or MPICH version 1.2.7p1
License: MIT license
Any restrictions for use by non-academics: License required
List of abbreviations
Genome-Wide Association Study
Single Nucleotide Polymorphism
Rapid Association Test
Standard Permutation Test
Markov-chain Monte Carlo
Han Chinese Beijing
This study was supported by the "Next-Generation Integrated Living Matter Simulation," a national project of the Ministry of Education, Culture, Sports, Science and Technology (MEXT), and by the Research Project for Personalized Medicine (MEXT).
- Hirschhorn JN, Daly MJ: Genome-wide association studies for common diseases and complex traits. Nat Rev Genet. 2005, 6 (2): 95-108.View ArticlePubMed
- Ozaki K, Ohnishi Y, Iida A, Sekine A, Yamada R, Tsunoda T, Sato H, Hori M, Nakamura Y, Tanaka T: Functional SNPs in the lymphotoxin-alpha gene that are associated with susceptibility to myocardial infarction. Nat Genet. 2002, 32 (4): 650-654. 10.1038/ng1047.View ArticlePubMed
- Onouchi Y, Gunji T, Burns JC, Shimizu C, Newburger JW, Yashiro M, Nakamura Y, Yanagawa H, Wakui K, Fukushima Y, Kishi F, Hamamoto K, Terai M, Sato Y, Ouchi K, Saji T, Nariai A, Kaburagi Y, Yoshikawa T, Suzuki K, Tanaka T, Nagai T, Cho H, Fujino A, Sekine A, Nakamichi R, Tsunoda T, Kawasaki T, Hata A: ITPKC functional polymorphism associated with Kawasaki disease susceptibility and formation of coronary artery aneurysms. Nat Genet. 2008, 40 (1): 35-42. 10.1038/ng.2007.59.PubMed CentralView ArticlePubMed
- Tokuhiro S, Yamada R, Chang X, Suzuki A, Kochi Y, Sawada T, Suzuki M, Nagasaki M, Ohtsuki M, Ono M, Furukawa H, Nagashima M, Yoshino S, Mabuchi A, Sekine A, Saito S, Takahashi A, Tsunoda T, Nakamura Y, Yamamoto K: An intronic SNP in a RUNX1 binding site of SLC22A4, encoding an organic cation transporter, is associated with rheumatoid arthritis. Nat Genet. 2003, 35 (4): 341-348. 10.1038/ng1267.View ArticlePubMed
- Misawa K, Kamatani N: ParaHaplo: A program package for haplotype-based whole-genome association study using parallel computing. Source Code Biol Med. 2009, 4 (1): 7-10.1186/1751-0473-4-7.PubMed CentralView ArticlePubMed
- Misawa K, Kamatani N: ParaHaplo 2.0: a program package for haplotype-estimation and haplotype-based whole-genome association study using parallel computing. Source Code Biol Med. 2010, 5 (1): 5-10.1186/1751-0473-5-5.PubMed CentralView ArticlePubMed
- Li Y, Willer C, Sanna S, Abecasis G: Genotype imputation. Annu Rev Genomics Hum Genet. 2009, 10: 387-406. 10.1146/annurev.genom.9.081307.164242.PubMed CentralView ArticlePubMed
- Marchini J, Howie B, Myers S, McVean G, Donnelly P: A new multipoint method for genome-wide association studies by imputation of genotypes. Nat Genet. 2007, 39 (7): 906-913. 10.1038/ng2088.View ArticlePubMed
- Nothnagel M, Ellinghaus D, Schreiber S, Krawczak M, Franke A: A comprehensive evaluation of SNP genotype imputation. Hum Genet. 2009, 125 (2): 163-171. 10.1007/s00439-008-0606-5.View ArticlePubMed
- Li Y, Abecasis GR: Mach 1.0: rapid haplotype reconstruction and missing genotype inference. Am J Hum Genet. 2006, 79: S2290.
- Culler DE, Gupta A, Singh JP: Parallel Computer Architecture: A Hardware/Software Approach. 1997, San Francisco, CA: Morgan Kaufmann Publishers
- The International HapMap Consortium: The International HapMap Project. Nature. 2003, 426 (6968): 789-796. 10.1038/nature02168.View Article
- Altshuler DM, Gibbs RA, Peltonen L, Dermitzakis E, Schaffner SF, Yu F, Bonnen PE, de Bakker PI, Deloukas P, Gabriel SB, Gwilliam R, Hunt S, Inouye M, Jia X, Palotie A, Parkin M, Whittaker P, Chang K, Hawes A, Lewis LR, Ren Y, Wheeler D, Muzny DM, Barnes C, Darvishi K, Hurles M, Korn JM, Kristiansson K, Lee C, McCarrol SA, et al: Integrating common and rare genetic variation in diverse human populations. Nature. 2010, 467 (7311): 52-58. 10.1038/nature09298.View ArticlePubMed
- Marchini J, Cutler D, Patterson N, Stephens M, Eskin E, Halperin E, Lin S, Qin ZS, Munro HM, Abecasis GR, Donnelly P: A comparison of phasing algorithms for trios and unrelated individuals. Am J Hum Genet. 2006, 78 (3): 437-450. 10.1086/500808.PubMed CentralView ArticlePubMed
- Nakamura Y: The BioBank Japan Project. Clin Adv Hematol Oncol. 2007, 5 (9): 696-697.PubMed
- Mardanov AV, Ravin NV, Kuznetsov BB, Samigullin TH, Antonov AS, Kolganova TV, Skyabin KG: Complete Sequence of the Duckweed (Lemna minor) Chloroplast Genome: Structural Organization and Phylogenetic Relationships to Other Angiosperms. J Mol Evol. 2008
- SNPHAP - A program for estimating frequencies of large haplotypes of SNPs. [http://www-gene.cimr.cam.ac.uk/clayton/software/]
- Rabenseifner R: Hybrid parallel programming on HPC platforms. The proceedings of the Fifth European Workshop on OpenMP, EWOMP 2003; Aachen, Germany. 2003
- Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, Higgins J, DeFelice M, Lochner A, Faggart M, Liu-Cordero SN, Rotimi C, Adeyemo A, Cooper R, Ward R, Lander ES, Daly MJ, Altshuler D: The structure of haplotype blocks in the human genome. Science. 2002, 296 (5576): 2225-2229. 10.1126/science.1069424.View ArticlePubMed
- Barrett JC, Fry B, Maller J, Daly MJ: Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005, 21 (2): 263-265. 10.1093/bioinformatics/bth457.View ArticlePubMed
- Kamatani Y, Matsuda K, Okada Y, Kubo M, Hosono N, Daigo Y, Nakamura Y, Kamatani N: Genome-wide association study of hematological and biochemical traits in a Japanese population. Nat Genet. 2010, 42 (3): 210-215. 10.1038/ng.531.View ArticlePubMed
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.