many confidence intervals can be classified as “recombination” after inspection of just two of the 101 points; see Figure 3.Figure 3Rapid classification of “recombination” variant pairs. This is a plot of 101 equally spaced D’ log-likelihoods for (rs58108140, rs140337953) in 1000 Genomes phase 1, used in Gabriel et al.’s method of identifying haplotype blocks. Whenever the upper end of the 90% confidence interval is smaller than 0.90 (i.e. the rightmost 11 likelihoods sum to less than 5% of the total), we have strong evidence for historical recombination between the two variants. After determining that L(D′=x) has only one extreme value in [0, 1] and that it’s between 0.39 and 0.40, confirming L(D′=0.90)<L(D′=0.40)/220 is enough to finish classifying the variant pair (due to monotonicity: L(D′=0.90)≥L(D′=0.91)≥…≥L(D′=1.00)); evaluation of the other 99 likelihoods is now skipped in this case. The dotted horizontal line is at L(D′=0.40)/220.Instead of saving the classification of every variant pair and looking up the resulting massive table at a later point, we just update a small number of “strong LD pairs within last k variants” and “recombination pairs within last k variants” counts while processing the data sequentially, saving only final haploblock candidates. This reduces the amount of time
