Crossover model of Lep-Rec reveals higher heritability of recombination

Abstract

Meiotic recombination, a process that reshuffles DNA between parental chromosomes, is almost universal in multicellular organisms. Recombination accelerates the response to selection by breaking the linkage and creating new allelic combinations that can affect the fitness of the progeny. This motivates us to characterise recombination rate variation and to take it into account in evolutionary models and studies. Linkage mapping utilises recombination to obtain recombination distances for genetic markers. With (whole) genome sequencing data, very dense linkage maps can be produced, locating crossovers precisely in the genome. This enables direct and accurate calculation of recombination distances, correction of possible errors in the genome and maps, and studying the relation between recombination and physical base-pair distances. This is now a relevant problem, as high-quality genomes are emerging for many species, and available dense linkage map data would supplement these genomes. Here we present a novel software Lep-Rec to compute the local re-combination rate, i.e. the percentage of crossovers per individual per megabase (cM/Mb) along the genome. Moreover, it can also estimate the underlying, only partly observed, tetrad crossover distribution for each chromosome, while modelling crossover and chromatid interference. Together with Lep-MAP3 and Lep-Anchor, Lep-Rec forms a complete toolbox for studying recombination and crossovers: Lep-MAP3 can robustly construct linkage maps for large number of markers and individuals, while Lep-Anchor can anchor, validate and correct genome assemblies using linkage maps, and together these software provide consistent and complete physical and linkage maps for further analysis with Lep-Rec. Lep-Rec is available from http://sourceforge.net/projects/lep-anchor. Finally, we demonstrate the performance of Lep-Rec using real and simulated data: It outperforms and simplifies currently available tools and its estimated crossover distribution can improve association analysis and heritability estimates of recombination.