ABC as a flexible framework to estimate demography over space and time: some cons, many pros

Abstract

The analysis of genetic variation to estimate demographic and historical parameters and to quantitatively compare alternative scenarios recently gained a powerful and flexible approach: the Approximate Bayesian Computation (ABC). The likelihood functions does not need to be theoretically specified, but posterior distributions can be approximated by simulation even assuming very complex population models including both natural and human-induced processes. Prior information can be easily incorporated and the quality of the results can be analysed with rather limited additional effort. ABC is not a statistical analysis per se, but rather a statistical framework and any specific application is a sort of hybrid between a simulation and a data-analysis study. Complete software packages performing the necessary steps under a set of models and for specific genetic markers are already available, but the flexibility of the method is better exploited combining different programs. Many questions relevant in ecology can be addressed using ABC, but adequate amount of time should be dedicated to decide among alternative options and to evaluate the results. In this paper we will describe and critically comment on the different steps of an ABC analysis, analyse some of the published applications of ABC and provide user guidelines.

MTML-msBayes: Approximate Bayesian comparative phylogeographic inference from multiple taxa and multiple loci with rate heterogeneity

“…To ascertain how well MTML-msBayes quantifies the congruence of divergence times under a number of different conditions, we conduct an extensive simulation analysis by generating PODS (pseudo-observed data sets; [14]) and quantifying the accuracy and precision of estimates on the known parameter values used to generate the PODS. Specifically, we assessed: 1.)…”

“…When this software pipeline is expanded to allow data consisting of large numbers of SNPs such that none of the information in the data are lost, we expect that a strategy involving genetic distances instead of Euclidian distances might work well for testing multi-taxa hypotheses or alternatively using the first four moments of sets of summary statistics calculated across loci and/or taxa [46,53]. For further information about the bourgeoning set of methods and strategies being developed for ABC, [12-14] provide thorough overviews.…”

“…By utilizing simulation and summary statistics to avoid the need to calculate an explicit likelihood function, ABC (approximate Bayesian computation) or "likelihood-free" methods can potentially tackle complex multi-taxa demographic models when more exact methods are not efficient [8]. Although some information in the data is sacrificed when only using summary statistics, ABC methods have been shown to compare well against methods that utilize an explicit likelihood function [9-11] by allowing efficient extraction of information from the data under explicit models that can be built from background information [12-14]. …”

“…"PODS"; [14]) in order to conduct simulation-based model validation as well as fine tune the ABC conditions with respect to choice of summary statistics and acceptance threshold. In addition, users can use available R scripts to conduct posterior model fitting in order to assess the fit of the simulation models to the observed data [13,14]. …”

Evolution under changing climates: climatic niche stasis despite rapid evolution in a non-native plant

“…Competing hypotheses about the introduction history of L. serriola in the two non-native groups were compared within an approximate Bayesian computing (ABC) framework using the software DIYABC [25,26]. The 14 scenarios (see electronic supplementary material, table S2) considered all four possible combinations of Europe and/or Asia as putative source regions, plus five scenarios in which at least one non-native group arose through admixture between Asia and Europe.…”