Effect of background selection on inferred recombination rates

To investigate the effect of background selection on our inferred recombination rates, we downloaded a genome-wide measure of background selection (B-statistics) from (77). B-statistics range in value from 0 to 1000 and reflect the relative loss in genetic diversity as a result of background selection, with 0 being a total loss in diversity and 1000 representing the truly neutral level of genetic diversity. The available B-statistics are reported in terms of coordinates on hg18, so we used LiftOver (76) to remap the coordinates to hg38. We took the data processed for analyzing determinants of recombination rate variation and further restricted to sites with reported B-statistics. We repeated the analyses of the effect of putative PRDM9 binding and chromatin state while controlling for background selection by including the B-statistics as linear covariates. The results are presented in fig. S5. While we observe a high correlation between the inferred recombination rate and B-statistics (Spearman’s ρ = 0.375, P < 2.2 × 10−16), the overall impact of chromatin state and PRDM9 binding remains comparable whether or not we control for B-statistics.

Note that B-statistics were originally computed by fitting distributions of selection coefficients for exonic and nonexonic regions to observed patterns of diversity (77). The impact of these distributions on the diversity at linked neutral sites depends on the genetic distance between the selected site and the neutral site, and hence requires knowledge of the fine-scale recombination map. Due to this circularity, it is difficult to determine whether the observed correlation between B-statistics and our inferred recombination rates is due to lower recombination rates directly causing higher levels of background selection (and hence lower inferred recombination rates being associated with lower B-statistics), or if higher levels of background selection result in a lower apparent effective population size, resulting in underestimated recombination rates. It may be possible to disentangle background selection from changes in local recombination rate by jointly inferring B-statistics and fine-scale recombination rates, but we leave such an undertaking for future work.

Note: The content above has been extracted from a research article, so it may not display correctly.



Q&A
Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.



We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.