MOESM1 of A method for allocating low-coverage sequencing resources by targeting haplotypes rather than individuals

Additional file 1: Figure S1. Expected haplotype imputation accuracy against the accumulated haplotype sequencing coverage, as estimated using a novel population-based imputation method (Battagin and Hickey, unpublished). A description of the prototype algorithm that was developed for the imputation of consensus haplotypes under the LCSeq approach and the simulated results on which the AlphaSeqOpt method is based. We generated 1x sequence data for the sires from a simulated population. The x-axis represents the expected accumulated coverage that each haplotype would receive. The y-axis represents the percentage of alleles phased and imputed for each haplotype. The imputation accuracy increased with the accumulated haplotype coverage until it plateaued. Haplotypes with a sequencing coverage of 10x accumulated from 20 individuals sequenced at 1x were imputed to the whole population with an accuracy of 0.88. Haplotypes with a sequencing coverage of 15x or 20x accumulated from 30 or 40 individuals sequenced at 1x were imputed to the whole population with an accuracy of 0.93 or 0.97, respectively. For accurate inference of a consensus haplotype, a certain amount of sequencing coverage must be accumulated. According to the results above, 10x or 15x could be good target coverages for the haplotypes