We selected the best localized of the two stronger signals and conducted a post hoc analysis of the false-positive signal (i.e., the signal centered over CDCA2 on chromosome 8) to determine whether there was a reasonable explanation for this strong false-positive finding obtained in the SOLAR analysis. Briefly, we found that this false-positive signal resulted from a chance deviation from random segregation between the A SNP at C8S775 in CDCA2 (the false-positive signal on chromosome 8) and the C SNP at C13S434 in FLT1 (a true signal on chromosome 13). Eighteen informative meioses involving these two SNPs were observed in the Yoruba kindred (pedigree 5) responsible for this false-positive signal (Figure 4). Under the hypothesis of linkage between the SNPs at C8S775 and C13S434, these meioses give rise to 15 nonrecombinants (NR) and 3 recombinants (R). Although this deviation from the null expectation of 9NR:9R is marginally significant by a one-sided Fisher’s exact test (p = 0.0375), it is not persuasive evidence in a linkage setting. However, when 50 marker-identical families are pooled, this minor nonrandom segregation is telescoped into
