4
$\begingroup$

I calculated an AMOVA from a genind object, with one hierarchical factor. In the table I obtain there are SSD values (for my grouping factor,"Error" and total) and sigma2 values (for my grouping factor and "Error")

I have two questions:

  • What does "error" stand for?
  • How do I calculate Fst? Which values do I have to use?

A genind object in R is an object which contains allelic information about a set of individuals.
This is the table that I get:

Analysis of Molecular Variance

Call: amova(formula = plantula.dist ~ g, nperm = 100)  

           SSD         MSD  df  
g     354992.9 354992.9297   1  
Error 310814.9    693.7834 448 
Total 665807.9   1482.8683 449 

Variance components:  
       sigma2 P.value  
g     1574.69       0  
Error  693.78         

Variance coefficients:  
       a   
224.9956   
$\endgroup$
5
  • $\begingroup$ Welcome to Bioinformatics! Could you please clarify what is a geneind object? And could you provide the exact output of the table? (If you provide how you get that table that would probably be also helpful to answer you) $\endgroup$
    – llrs
    Jun 8 '17 at 12:35
  • $\begingroup$ I tried to insert the table, though I can't quite format it right... sorry, I'm still figuring this out! $\endgroup$ Jun 8 '17 at 12:59
  • $\begingroup$ By Fst are you refering to Fixation index? $\endgroup$
    – llrs
    Jun 8 '17 at 13:05
  • $\begingroup$ Please also indicate the non-standard packages you are using. I’m assuming amova comes from ‹pegas›? $\endgroup$ Jun 8 '17 at 13:05
  • $\begingroup$ Yes @Llopis I mean fixation index, and yes Konard Rudolph, the amova I am using is from the package pegas. Sorry for not specifying! $\endgroup$ Jun 8 '17 at 13:19
1
$\begingroup$

Error is the estimation that can't be fitted by your parameters (in this case g). The amova function in pegas package does the following:

This function performs a hierarchical analysis of molecular variance as described in Excoffier et al. (1992). This implementation accepts any number of hierarchical levels

And as per the definition of Fixation index, you need the variance in the frequency of the allele between different subpopulations, weighted by the sizes of the subpopulations, the variance of the allelic state in the total population. Which is not provided by amova. Thus you cannot compute the fixation index from amova results.

$\endgroup$
2
  • $\begingroup$ That's what I thought. Thank you for confirming it. And thanks for the explanation for error! $\endgroup$ Jun 10 '17 at 9:19
  • $\begingroup$ @BeatriceBaldi if this answers your question, you can mark as solved by checking the tick at the left of the answer. This way you let others know that this problem is solved $\endgroup$
    – llrs
    Jun 10 '17 at 10:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.