We used a sample of 1,295 two-generation families with multiple offspring to study the recombination landscape and the genetic basis of meiotic recombination. Our analysis showed that the locations of the recombination events differ across the genome; most of the crossovers occur at the ends of chromosomes. In particular, 70% of male recombination jungles are located in the 5% most telomeric parts of chromosomes. This pattern is observed in samples from the AGRE and FHS collections and also in previous studies of CEPH and Icelandic populations. We also found extensive individual variation in the number of recombination events per meiosis in both females and males. To determine the genetic basis of this variation, we carried out genome-wide association analysis. We treated the number of recombination events as quantitative traits to map the genetic variants that influence recombination phenotype. We found three loci that influence the female recombination phenotype, and three loci that influence the male recombination phenotype.
Our GWAS analysis replicated the previous findings that an inversion on chromosome 17q21.31 is associated with higher female recombination rates, and that variants in RNF212
influence recombination rates in males. The previous work showed association between haplotypes in RNF212
and male and female recombination, with opposite effects in the two sexes 
. Specifically, they found SNP rs1670533 associated with female recombination and rs3796619 associated with male recombination. Neither of these SNPs was in the set we examined, but the linkage disequilibrium in this region is very high, and our dataset does contain SNPs that are perfect surrogates (r2
1.0) in HapMap for each of these. Our surrogates for the Icelandic male SNP include rs4045481 and rs11939380, both of which had P-values on the order of 10−6
for male recombination in our combined dataset. Our surrogates for the Icelandic female SNP are rs6827357 and rs20114318. Both of these showed P-values on the order of 0.01 for females in our combined dataset. As in the Icelandic population, we observed opposite effects on male and female recombination at these SNPs and throughout the haplotype block. While the P-values for female recombination in our dataset fall far short of genome-wide significance, they do show a weak association and are quite consistent with the Icelandic results. It is particularly notable that this is the first replication of the curious opposite effects on male and female recombination previously observed.
In addition to these known loci, we uncovered additional polymorphic regions that are associated with recombination phenotypes. In females, we found variants on chromosome 10 near a poorly characterized gene, KIAA1462, and those on chromosome 1 near PDZK1 to be associated with recombination rates. These variants along with those on chromosome 17q explain approximately 6% of the total variation in female recombination phenotype in the AGRE and FHS samples. In males, we identified variants on chromosome 9 near UGCG and on chromosome 7 near NUB1 to be associated with recombination phenotype. In the mouse, the expression of Ugcg and Nub1 in prophase I further supports their potential roles during meiosis. The variants in RNF212 and those on chromosomes 7 and 9 explain about 5.4% of variation in male recombination.
Results from our genetic mapping study enhanced our understanding of meiotic recombination. It appears that gender differences in recombination rates and pattern result from differences in the regulation of female and male meiosis. Our genetic mapping results showed that DNA variants in different genes are associated with female and male recombination phenotypes; we did not find any variants that are significantly associated with both female and male recombination phenotypes. Second, we identified multiple unlinked SNPs that are associated with recombination phenotypes suggesting that multiple polymorphic regulators influence these phenotypes. This likely provides a mechanism for variability in recombination which is essential for genetic diversity while maintaining the number of recombination events within a range that ensures proper disjunction. Each of the variants that we identified explains only a small fraction of the individual variation in recombination. Together the three loci that contribute to female recombination explain less than 10% of the variation, and the same for male recombination. Unlike most essential cellular processes, recombination events must differ between individuals to maintain genetic diversity. However, the system cannot be so flexible that it fails to ensure proper segregation of chromosomes. Having many regulatory steps achieves the goal of allowing some range of events to occur while ensuring that the number of recombination events does not deviate too much to cause improper chromosome segregation or non-disjunction. Although, we have identified six variants that influence recombination events, we expect other variants still need to be identified. Characterization of genetic variants that influence natural variation in meiotic recombination will allow a better understanding of normal meiotic events as well as non-disjunctions which lead to chromosomal abnormalities, the primary cause of miscarriages.