EMINIM is a software for imputation of unobserved genotypes using a set of reference haplotype panel at a higher-density SNP set such as HapMap, and lower-density genotypes of a target individual using such as genotyping arrays. Our method is adaptive in the sense that it learns transtition/emission parameters of hidden Markov model using EM algorithm with standard Poisson point process for modeling recombination. Our method is memory-efficient by efficiently utilizing silent states in the HMM. EMINIM provides interfaces both for human and inbred model organisms such as laboratory mice. Our method is demonstrated to have ten times smaller errors than the previous imputation method designed for inbred mouse for high-confidence genotypes.
The current implementation of EMINIM is under beta status and is available for download. The official release should be available any time soon. If you have you any further question, please contact Hyun Min Kang.
This web-site is based upon work supported by the National Science Foundation under Grant No. 0513612, No. 0731455, and No. 0729049, and National Institutes of Health under Grant No. 1K25HL080069. This research was supported in part by the UCLA subcontract of contract N01-ES-45530 from the National Toxicology Program/National Institute of Environmental Health Sciences to Perlegen Sciences. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.