Haplotype estimation in polyploids using DNA sequence data

Polyploid organisms possess more than two copies of their core genome and therefore contain k>2 haplotypes for each set of ordered genomic variants. Polyploidy occurs often within the plant kingdom, among others in important corps such as potato (k=4) and wheat (k=6). Current sequencing technologies enable us to read the DNA and detect genomic variants, but cannot distinguish between the copies of the genome, each inherited from one of the parents. To detect inheritance patterns in populations, it is necessary to know the haplotypes, as alleles that are in linkage over the same chromosome tend to be inherited together. In this work, we develop mathematical optimisation algorithms to indirectly estimate haplotypes by looking into overlaps between the sequence reads of an individual, as well as into the expected inheritance of the alleles in a population. These algorithm deal with sequencing errors and random variations in the counts of reads observed from each haplotype. These methods are therefore of high importance for studying the genetics of polyploid crops.