See page 129 of the Arlequin manual below

http://cmpg.unibe.ch/software/arlequin35/man/Arlequin35.pdf

where they suggest to "see Michalakis and Excoffier, 1996, as well as Rousset, 1996 , for details on the relationship between FST and RST" 

My sincere thanks for pointing this out, I looked through the manual before but did not catch that!

Hi Cemal

You need to recode your SSRs in terms of the number of repeats!! An then tick the appropriate boxes in Arlequin

Hello,

I have to do the same.. I have haplotype frequencies and loci with alleles and have to estimate Rst with Arlequin.. any help?

Hi Irene

I am fond to help people, but what do you expect from my help? I suggest you read the manual, explain us what you did, what the failures are and then, what are your specific needs.

Any help? is just too vague

Cheers

Pablo

Dear Irene, if you are new to genetic distance estimations, you may want to start with yhrd online amova/mds tool, where you can get a paiwisr matrix of fst, rst, etc values for your new dataset and any other selected datasets that you chose from yhrd and/or upload with your new data. Yhrd online tool is not perfect, but a good place to start... you can also use arlequin or even spagedi which are all freeware, but preparing input files might be a bit of a challenge, perhaps an easy task with you bioinf background. See my papers on turkish cypriots as an example (ann hum biol) or northern iraqi populations (plose one). In the latter paper you will also find an alternative method based on nei’s da distance, etc. hope this helps

Pablo and Cemal thank you for your answers.

Is important to calculate those values, to continue with PCA plots and contour maps.

Well I tried yhrd amova, the first time it made the calculations, on the second time there was an error, maybe the data in the file.

I have 2 files of excel, the one with haplotype frequencies of many populations and I want to calculate the Rst and Fst. I tried to convert my excel to .arp to run in Arlequin (using PGDSpider_2.1.1.5) and then I put my file into Arlequin, choose what calculations I want but never runs.

I want to know, is there a specific format that Arlequin read the data?

For example, is there a specific beginning and ending word/ point/ key word etc?

I will see your papers Cemal.

Irene,

Both the YHRD online AMOVA/MDS tool and Arlequin, etc are of course very sensitive to data formating issues. I think at least for YHRD, you can try using the data validator tool first... also check their examplar input file provided. With Arlequin and Spagedi, etc, I have a close collaborator who prepares input files via Phython scripts, so we longer have problems like you describe. For Arlequin as well, please check user guides for the details on input file formats, or please ask more specific questions : )

Cemal,

I read carefully the manual of Arlequin for the input files.

I create one with the format it said but when I go to run it in Arlequin, it says "0.05 is not a valid floating point".

Hi Irene

Could it be that your locale settings require a colon (,) instead of a point (.) to express decimal numbers?

Pablo

Hello, to start with Rst calculations might be a bit problematic with Y-STRs for a number of reasons... To start with, all those haplotypes with (a) null alleles, (b) partial alleles (e.g. DYS458*18.2, etc), and (c) multi-allelic patterns (e.g. allelic duplications), etc have to be removed prior to analyses. This is something automatically done with the YHRD AMOVA/MDS tool, and one will have to to do something parallel with Arlequin. This would still be fine with some population datasets, but consider a population with a higher frequency of partial alleles, say Semitic populations with the DYS458*.2 variants, which mostly correlate with te J1 haplogroup, and which may make up to nearly 50% of the given population if not more. Do you really want to cut the population dataset in half or even more prior to the Rst analysis, even if this would mean cutting out perhaps the most relevant haplogroup for such populations? Alternatively one can remove the entire DYS458 locus which would be a more conservative approach, but which would in turn reduce the data resolution. As a result of problems like these and others, we moved to the use of Spagedi software for Rst calculations at one point, and ultimaly we now normally prefer to do allele frequency-based Nei's DA distance calculations instead of Rsts. In our hands, Nei's DA distance based estimate and visualisation via phylogenegtic trees and MDS plot seem to yield more reliable results with Y-STR datasets, though we still do calculate Rst at times, and only after cutting down our datasets to reduce above explained 'Rst unfriendly data'