Haplotype resolved genomes: Computational challenges and applications

Genomes of diploid organisms, like humans, are organized in pairs of chromosomes, one inherited from the father and one from the mother. Each homologous chromosome harbors a specific set of parental alleles, called haplotype. Unfortunately, to obtain haplotype information using current methods remains challenging. Here we introduce a single cell DNA template strand sequencing (Strand-seq) as a novel haplotyping approach able to separate parental alleles along the entire length of all chromosomes. We demonstrate this by building a complete haplotypes for HapMap individual (NA12878) at high accuracy (concordance 99.3%), without using generational information or statistical inference. Furthermore we mapped all meiotic recombination events in a family trio with high resolution (median range ~14 kb), and phased larger structural variants like deletions, indels as well as balanced rearrangements like inversions. The single cell resolution of Strand-seq allowed us to observe loss of heterozygosity regions in a small number of cells, a significant advantage for studies of heterogeneous cell populations, such as cancer cells. Lastly, we prove that integration of Strand-seq with other whole-genome sequencing methods brings significant increase in haplotype completeness while reducing sequencing costs. The implementation of Strand-seq and our analysis pipeline brings a powerful, high-throughput approach to assemble haplotypes that will open up new possibilities to study diploid architecture of human genomes in health and disease.