Measuring population-level plant gene flow with topological data analysis

Abstract

Hybridization is an important factor influencing the evolution of many plant species. Measurement of interspecific gene flow and quantification of recombination provide insight into the processes that shape population genetic diversity. Recently, a new approach called topological data analysis (TDA) has begun to be implemented to track genomic recombination between species and populations. However, existing TDA methods tend to underestimate gene flow in cross-population models. Here, we present a novel graph-based approach to measuring gene flow, which occurs during ongoing speciation and hybridization in plants in various mating systems. We combine minimal spanning networks with topological filtering and then test the resulting parameter measurements on simulated primary divergence and secondary contact processes under different mating conditions. The resulting parameters, based on the first Betti number (total number of one-dimensional cycles), showed a high positive correlation with the number of migrants (Nm) in almost all cases of secondary contact. Considering the modularity of MSN networks, we have devised a parameter that works well in the event of primary divergence, where gene flow is at its highest level. This approach is suitable for various types of data, including SNPs, microsatellites, and binary restriction fragment data.