SPAGEDi

SPAGeDi (Spatial Pattern Analysis of Genetic Diversity) is a computer package primarily designed to characterise the spatial genetic structure of mapped individuals and/or mapped populations using genotype data of any ploidy level. It can compute various statistics describing relatedness or differentiation between individuals or populations by pairwise comparisons, and analyse how these values are related to geographical distances, 1°) in a way similar to a spatial autocorrelation analysis, 2°) by linear regressions (the slopes of these regressions can be used to obtain indirect estimates of gene dispersal distances parameters such as neighbourhood size). SPAGeDi can also treat data without spatial information, providing global estimates of genetic differentiation and/or matrices of pairwise statistics between individuals or populations. Data from dominant markers such as AFLP or RAPD can also be treated to estimate pairwise kinship or relationship coefficients between individuals (Hardy 2003 ME.pdf). The statistics computed include Fst, Rst, Nst, Ds (Nei’s standard genetic distance), and (delta mu)2 (Goldstein and Pollok 1997) for analyses at the population level and, for analyses at the individual level, pairwise kinship, relatedness and fraternity coefficients (with different estimators for each) as well as Rousset’s distance between individuals and a kinship analogue based on allele size. Jackknife over loci gives approximate standard errors. Permutations of locations, individuals or genes provide ad hoc tests of spatial structure, population differentiation or inbreeding, respectively. A new allele size permutation test also allows to check whether microsatellite allele sizes carry a relevant information about genetic structure (Hardy & al. 2003 Genetics.pdf). In addition, the actual variance of the statistics can be estimated following the method of Ritland (2000), providing a measure necessary for marker-based inference of the heritability or Qst of quantitative traits.