Skip to main content

Table 1 Brief background information about the programs and methods used in the present study

From: Imputation of missing genotypes within LD-blocks relying on the basic coalescent and beyond: consideration of population growth and structure

Program/Method Background information Application to the present study Reference
BATWING The program reads multi-locus haplotype data and uses a Markov chain Monte Carlo method based on coalescent theory to generate approximate random samples of the underlying gene genealogy. BATWING allows specification of the population growth and structure models with their corresponding prior distributions. Estimation of gene genealogies underlying haplotypes under the basic coalescent and also considering population growth and structure [12]
Genetree The program constructs gene trees describing the history of a sample of DNA sequences and calculates maximum likelihood estimates of the time to the most recent common ancestor and mutation, migration and growth rates, also in substructured populations. Exclusion of incompatible sites by pairwise four-gamete tests [11]
IMPUTE2 Computer program for phasing observed genotypes and imputing missing genotypes. Basically, phasing and imputation are alternatively iterated in a Markov chain Monte Carlo framework which accounts for phase uncertainty. Used as gold standard for genotype imputation assuming no recombination and also considering regional recombination rates [3]
msms Extension of Hudson’s coalescent simulator ms, which also permits to study selection. Since selection was not considered in the present study, our haplotypes were simulated using standard coalescent methods: genealogies were generated by tracing randomly sampled alleles backwards in time. Haplotype simulation under the basic coalescent and also considering population growth and structure [8]
SHAPEIT2 Fast and accurate method for phasing from genotype or sequencing data. Phasing of real genotype data [10]
SumTrees The program constructs a summary tree based on tree samples provided by the user. Supported methods for summary tree construction include the Maximum Clade Credibility Topology, and the majority-rule clade consensus. Combination of gene genealogies estimated by BATWING into a majority-rule consensus tree [14]