First, we run a fast O(MN)-time scan to find probable IBD with other haploid chromosomes (according to phase calls made in step 1). This procedure begins analogously to the first component of step 1; again, we look for long segments of IBS>0 (now between the diploid proband and haploid potential surrogates), now allowing a single mismatch site (IBS=0) within runs. We then attempt to extend the identified seed matches and record the ten longest matches covering each SNP block (as defined above).
