The yield of LRP presented here is high, but it would likely be even higher if not for the current algorithm, which is a first attempt at implementing a procedure that can process data in mass and give reliable results, and is both conservative and inefficient. Much information is not utilized, and we expect that by incorporating a number of refinements (Supplementary Note) to the procedure, the yield would increase without elevating the error rate. Currently, yield is limited by the criterion that an individual is considered a surrogate parent only if he is IBS ≥ 1 with the proband for the entire extended region. This could not only rule out true surrogate parents due to a single genotyping error, but also eliminates individuals who share a very long haplotype with the proband, part of it extending beyond the extended region on one side, but the shared haplotype does not cover the entire target or extended region. Utilizing these individuals in a proper manner is crucial to our ultimate goal of phasing chromosomes in their entirety, achievable by stitching together phasing results from overlapping target regions.
