However such methods face several limitations; Lander-Green algorithm based approaches have computational and space complexity that scale exponentially with sample size; they can be sensitive to genotyping error and they can only phase sites where at least one member of the pedigree is not heterozygous. The last point is particularly crucial, as it means the haplotypes will not be ‘complete’ and cannot be easily used in pre-phasing and imputation which is now a standard part GWAS pipelines [11]. If founders in these pedigrees have been sequenced with the aim of imputing sequenced variants from founders into descendants who have been assayed on microarrays, then a pedigree phasing method that overcomes these issues will be especially useful.
