SpaceMix, and a brief history of Spatial Genetics

Incorporating spatial data to inform studies of the population demography of a species has a long history of interest. From inferring geographical clines in Principal Components Analyses (Menozzi et al. 1978), using location data as “informative priors” during model-based estimation of admixture (Hubisz et al. 2009), using phylogenetic trees (and other distance based methods) and superimposing them upon geographical distributions to make predictions about what has come to be known as “phylogeography” of a species, measuring the correlation between geographic distance matrices and genetic distance matrices (sensu Peakall and Smouse 1999), estimating spatial autocorrelation (eg. Moran’s I, correllograms – see Brian Epperson’s book for an excellent review, also Sokal and Oden 1991) to discover directional clines in the genetic-spatial distribution of a species, pruning variance-covariance matrices in genetic data using graph-theoretical/network algorithms to discover geographical-genetic structure, detecting differences in allele frequency spectra of populations to detect founder effects and range expansions (see Peter and Slatkin 2013), just to name a few.


Inferred map of human admixture using SpaceMix from Bradburd et al. (2015).

At the core of all these methods is the variance-covariance structure in the genetics (primarily observed and/or ancestral allele frequency distribution), and the apparent geographical distribution of the species.
A recent method, developed by Gideon Bradburd and colleagues, called SpaceMix makes the visualization of “geo-genetic” admixture a lot more intuitive, and inferential (as against a lot of methods mentioned above which involve quite a bit of “hand-waving” to build a story out of the data). To wit, SpaceMix utilizes a Bayesian MCMC framework to estimate the parameters such as genetic covariance and admixture proportions in a “geo-genetic” space.

The result is a “geogenetic” map in which the distances between populations are effective distances, indicative of the way that populations perceive the distances between themselves: the “organism’s-eye view” of the world.

The authors utilize several simulated and real datasets to describe the utility of SpaceMix, and while still in beta, we really look forward to using/reviewing the code in the coming months!
Menozzi, Paolo, Alberto Piazza, and L. Cavalli-Sforza. “Synthetic maps of human gene frequencies in Europeans.” Science 201.4358 (1978): 786-792.
Hubisz, Melissa J., et al. “Inferring weak population structure with the assistance of sample group information.” Molecular ecology resources 9.5 (2009): 1322-1332.
Avise, John C., et al. “Intraspecific phylogeography: the mitochondrial DNA bridge between population genetics and systematics.” Annual review of ecology and systematics (1987): 489-522.
Smouse, Peter E., and Rod Peakall. “Spatial autocorrelation analysis of individual multiallele and multilocus genetic structure.” Heredity 82.5 (1999): 561-573.
Excoffier, Laurent, Peter E. Smouse, and Joseph M. Quattro. “Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data.” Genetics 131.2 (1992): 479-491.
Sokal, Robert R., and Neal L. Oden. “Spatial autocorrelation analysis as an inferential tool in population genetics.” American Naturalist (1991): 518-521.
Epperson, Bryan K. Geographical Genetics (MPB-38). Princeton University Press, 2003.
Dyer, Rodney J., and John D. Nason. “Population graphs: the graph theoretic shape of genetic structure.” Molecular Ecology 13.7 (2004): 1713-1727.
Peter, Benjamin M., and Montgomery Slatkin. “Detecting range expansions from genetic data.” Evolution 67.11 (2013): 3274-3289.
Bradburd, Gideon, Peter L. Ralph, and Graham Coop. “A Spatial Framework for Understanding Population Structure and Admixture.” bioRxiv (2015): 013474.

This entry was posted in bioinformatics, howto, population genetics, R, software and tagged , , , . Bookmark the permalink.