Mathematical function for predicting recombination from map distance

Abstract

Map distance, measured in centimorgans, is used in genetics to represent the average number of crossovers between two loci in a gamete and can be translated to recombination frequency with mapping functions. Commonly used mapping functions, such as Haldane or Kosambi functions, are single-variable functions and yield one recombination prediction per map distance. Here we show that these approaches contain systematic error and formulate a new function to predict recombination frequency from the map distance. Our new function is paradigm shifting as it takes into account the variation in the number of crossovers and their localisation, yielding context-specific estimates of recombination frequency that can vary for a given map distance. The function is chromosome-specific, making it generalizable for different sexes and organisms. We validate the new approach using empirical data from stickleback fishes and humans and show that it outperforms the conventional mapping functions and produces more accurate predictions about recombination frequency between loci.