(c) Rémy Petit (petit_at_pierroton.inra.fr), INRA-Bordeaux
Petit, El Mousadik & Pons. 1998. Identifying populations for conservation on the basis of genetic markers. Conservation Biology12, 844-855.
This program provides a measure of the contribution of each population to total diversity (measured by H, the expected heterozygosity, or by R, the allelic richness). This contribution is splitted in two components: one due to the diversity of the population, the other due to its differentiation from the remaining populations. The program may also be used to obtain allelic richness after rarefaction (to a sample size chosen by the investigator) for a set of populations. It can be used in conjunction with the program haplodiv, based on the paper by Pons & Petit 1995, TAG 90, 462-470, which will provide standard errors for the diversity and differentiation parameters. The input file is a text file (see example: rartest.txt), where the first line indicates the number of haplotypes (here it is 18), the number of populations (here it is 4), and the rarefaction size (it should not be larger than the smallest population sample size; here the rarefaction size is 10, and the smallest sample size is 20). Then follows the data for each population (line), with the number of each haplotype in each population (don't use relative frequencies):
18 4 10
1 0 1 0 0 0 1 1 ...(18 columns)
0 1 2 1 1 0 13 0 ...
0 0 8 0 0 3 6 0 ...
1 0 9 0 0 3 7 1
Results can be seen in the output file (rartest.out here; to be printed horizontally).
General measures are given first: within population diversity (Hs), total diversity (Ht), and Gst are given, followed by similar measures based on allelic richness. Then you get the results for each population : H, its standard error, allelic richness after rarefaction, the divergence from the other populations (DHs, DHt, DGst, see the paper in Conservation Biology), and the contributions Ct, Cs, Cd followed by the contributions for allelic richness measures. The program is written for an haploid gene but may be used for nuclear genes, assuming Hardy-Weinberg equilibrium. How to proceed when there are several loci? Do not take the mean across Gst or across Contributions. They are ratios, so you should take the mean of the numerator and the mean of the denominator separately. For the denominator: take the mean of hT and Rt-1 across loci. For the numerator: multiply the contributions by hT or Rt-1 (respectively for contributions to diversity or to allelic richness) and take the mean of these products across loci. Then compute the ratio of the two means. Good luck!
*Version 1.02: 03/2006. The bug found by Juha-Pekka Vähä (Biology University of Turku, Finland) was fixed. Depending on the input, results for Cpt, Crs, Crd and RST could be wrong.
*Version 1.01:01/2006. The program doesn't bug when the number of population > number of cytotype.
*Version 1.0: 04/2005. No More Parameters Limitation: New Graphical Interface. Delphi Program.
*First Version: September 1999. Parameters limitation size: the number of haplotypes limited to 45, the number of populations limited to 30. Turbo Pascal Program
- Logiciel CONTRIB (ZIP)