[adegenet-forum] interpretation of DAPC axes

[Jombart, Thibaut]

Hello, 

DAPC scatterplots can be interpreted as PCA and PCoA scatterplots, only DAPC provides an optimal visualisation of differences between clusters.
I wouldn't be too concerned about a horseshoe effect for the DAPC examples. For instance, in the case of the IBD models, only the first dimension should be interpreted.

The figure you describe sounds a bit surprising, and could probably be created by different processes, but hierarchical clustering is not what I would think of first (maybe more 2-dimensional IBD). Can you post the corresponding figure?

Best

Thibaut
________________________________________
From: adegenet-forum-bounces at lists.r-forge.r-project.org [adegenet-forum-bounces at lists.r-forge.r-project.org] On Behalf Of Pfau, Dr. Russell S. [PFAU at tarleton.edu]
Sent: 08 December 2010 16:45
To: adegenet-forum at lists.r-forge.r-project.org
Subject: Re: [adegenet-forum] interpretation of DAPC axes

I apologize, there was a type that may make my situation difficult to understand.  The next to last sentence should read as "A PcoA and STRUCTURE analysis of the same data *show* hierarchical clustering (two divergent populations, one having subpopulations with introgression occurring among the subpopulations)".

Apologies,
Russell

>|-----Original Message-----
>|From: adegenet-forum-bounces at lists.r-forge.r-project.org
>|[mailto:adegenet-forum-bounces at lists.r-forge.r-project.org] On Behalf
>|Of Pfau, Dr. Russell S.
>|Sent: Wednesday, December 08, 2010 10:41 AM
>|To: adegenet-forum at lists.r-forge.r-project.org
>|Subject: [adegenet-forum] interpretation of DAPC axes
>|
>|I'm a little uncertain how to interpret the axes of the scatterplot of
>|the DAPC.  I know how to interpret the axes of a PCA or PcOA.  For
>|example, Fig 8 of your BMC Genetics paper shows a U-shaped
>|pattern...which is also seen in Fig 4 c and d.  Is this the 'arch
>|effect' or 'horseshoe effect'?  Is this a problem?  How is it
>|interpreted?
>|
>|I've just run a DAPC of one of my datasets (with 24 populations) and
>|the DAPC scatterplot shows populations arranged in a *prefect*
>|circle...a donut shape.  How would one interpret that?  A PcoA and
>|STRUCTURE analysis of the same data who hierarchical clustering (two
>|divergent populations, one having subpopulations with introgression
>|occurring among the subpopulations).  The donut makes me think this is
>|a ring species which I know it is not.
>|
>|Thanks!!
>|Russell
>|_______________________________________________
>|adegenet-forum mailing list
>|adegenet-forum at lists.r-forge.r-project.org
>|https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/adegenet-
>|forum
>|
>|No virus found in this incoming message.
>|Checked by AVG - www.avg.com
>|Version: 8.5.449 / Virus Database: 271.1.1/3302 - Release Date:
>|12/08/10 07:34:00
_______________________________________________
adegenet-forum mailing list
adegenet-forum at lists.r-forge.r-project.org
https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/adegenet-forum