[en] Haplotype reconstruction is important in many applications in animal genomics. In livestock species, thanks to the availability of large half-sibs families and genotyped relatives, phasing methods can rely on strong familial information and results in families with more than 10 offspring are very accurate. However, most methods are sensitive to genotyping and map errors which will be more common with next generation sequencing data. Such problems are particularly important when studying recombination rate as we plan to do in the near future. We
herein describe a novel algorithm which is robust to genotyping errors and which can identify errors in marker maps. Using a large dairy cattle data set genotyped with high-density genotyping arrays, we show that the novel algorithm strongly reduces the occurrence of spurious
cross-overs due to different sources of errors, and identifies map errors for most of the bovine autosomes. The implemented version is still experimental and further research will be conducted to characterize the novel method (including simulations) and to fully describe the identified map errors.