Softwares like pop gene and Genalix can be use to calculate them

Alleles per locus you can just count (i.e. the number of different alleles for a locus in a population), while effective number of alleles is corrected for different sample size in different populations. You can e.g. use FSTAT to calculate it.

Effective number of alleles (Ae) is not corrected for different sample sizes (this is allelic richness Ar). The effective number of alleles is the number of alleles with equal frequencies (p1=p2=p3=..) it would take to give you the same He as in your study population (where the allele frequencies are not equal). Alleles with low frequencies contribute very little to the effective number of alleles so a large discrepancy between 'number of alleles' and 'effective number of alleles' tells you that you have quite a few low frequency alleles in the population.

Effective number of alleles (Ae) is a reciprocal of expected homozygosity, so it is absolutely correlated with genetic diversity measured as the expected heterozygosity (or a sum of squares of allele frequencies at a locus). One consequence of being a reciprocal function is that Ae needs to be averaged across loci using harmonic mean instead of arithmetic mean...

Indeed, I have mixed up effective number of alleles with allelic richness in my answer; I hadn't read the question properly!

Igor comment is absolutely right.

But just to clarify it, you can estimate Ae per locus. Now, if you want an average estimate across loci, you should do it with the harmonic mean, as Igor says.

GenAlex option Frequency.. and select Het by Pop. There you get a column named Ne, thats your number of effective alleles. While Na is your number of alleles.

Yes, you can use GenAlex by selecting Het.

Tiziana A Gelmi Candusso and Markus Ruhsam I am two years late to this conversation and new to GenAlEx. Can you clarify for me please that Na (number of alleles) as it is calculated in GenAlEx is the same as allelic richness? Does it use rarefaction and thus deal with slightly inequal sample sizes (mine range from 28 individuals to 33). My thanks. Erin

Hi Erin,

Na in Genalex is the number of different alleles averaged over all loci and is not corrected for sample size. Genalex doesn't do allelic richness (the one that takes unequal sample size into account) but the R package diveRsity for example does. However, if your sample sizes range between 28 and 33 I don't think allelic richness will be very different from Na.

Markus Ruhsam thank you so much for clearing this up for me! Much appreciated. I will use diveRsity to calculate allelic richness then, and will compare with the Na findings from GenAlEx. Cheers, Erin

Hi everyone,

Is it possible to calculate Na (number of different alleles averaged over all loci) in R? Someone could indicate some package for that?

Hi Flávia Rogério

You can calculate the mean number of alleles per locus (A) and an effective number of alleles per locus (Ae) in genetic diversity analysis by POPGENE 32, Microsoft Window-based freeware for population genetic analysis, Version 1.32.

Good-by

The R package 'diveRsity' can calculate Na; https://cran.r-project.org/web/packages/diveRsity/index.html