Minimum spanning network for population, included diploid and triploid individuals

[Daniel Kulik]

Hello,
I want to do minimum spanning network for population (base on Bruvo's distance), included diploid and triploid individuals (alloploidy). The problem is that when I create genind object, I can choose only on ploidy. As far as I understand from Bruvo's paper (2004), command bruvo.msn will not care about that I put information about ploidy when I was creating genind object?  It will take differences between loci and even in triploid I have loci with two alleles. So I create genind object where I have data for diploids and triploids and set up as triploids genotypes. Then, I performed analysis by bruvo.msn, I got MSN tree but still I'm not sure if I did correctly?

Thank you in advance,

Daniel

[Zhian Kamvar]

This is correct. Bruvo's distance will apply one of the four models to fill in the missing alleles from your data. I've written about it in the "Polyploids" section in https://grunwaldlab.github.io/poppr/reference/bruvo.dist.html#details.

The way to prepare the data for poppr is and has been to use an allele labelled as "0" to pad all the genotypes so that the genind object can hold the data. Bruvo's distance knows to treat the "0" allele as missing (though it makes no difference between null and missing alleles). 

If you want to do analyses other than Bruvo's distance that involve allele counts, you should use the recode_polyploids() function to strip those alleles from the genind object (see: https://grunwaldlab.github.io/poppr/reference/recode_polyploids.html)

I hope that helps.

Best,

Zhian

--
You received this message because you are subscribed to the Google Groups "poppr" group.
To unsubscribe from this group and stop receiving emails from it, send an email to poppr+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/poppr/bf218833-592e-4aec-9d44-5e634607bd68n%40googlegroups.com.