<div dir="ltr">Hi there, <div><br></div><div>I would definitely second Mark's comment and use cross-validation here. </div><div><br></div><div>Also for the clustering, I would give snapclust a try - I have just pushed a new version on github which is now properly documented. Especially check what the 'optimal k' is according to the various goodness of fit stats (snapclust.choose.k) - AIC, AICc, BIC, KIC.</div><div><br></div><div>Best</div><div>Thibaut</div></div><div class="gmail_extra"><br clear="all"><div><div class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><br>--<br>Dr Thibaut Jombart<br>Lecturer, Department of Infectious Disease Epidemiology, Imperial College London<br>Head of RECON: <a href="http://repidemicsconsortium.org" target="_blank">repidemicsconsortium.org</a><br>WHO Consultant - outbreak analysis</div><div><a href="https://thibautjombart.netlify.com" target="_blank">https://thibautjombart.netlify.com</a><br>Twitter: @TeebzR<br>+44(0)20 7594 3658</div></div></div></div>
<br><div class="gmail_quote">On 1 February 2018 at 20:36, Mark Coulson <span dir="ltr"><<a href="mailto:Mark.Coulson.ic@uhi.ac.uk" target="_blank">Mark.Coulson.ic@uhi.ac.uk</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div lang="EN-GB" link="#0563C1" vlink="#954F72">
<div class="m_-9004693243091395469WordSection1">
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d">Hi Nikki,<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d">Your interpretation of the plot seems correct, however I’d ask if you ran the xvalDAPC cross validation? It may be that you have kept
too many PCs so are overfitting the data. The xvalDAPC will find the optimal number of PCs to retain for your two groups. Then use this number of PCs to run a new DAPC. It will likely result in more overlap between the two groups, which would then be more
consistent with the low differentiation you are seeing based on FST. <u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d">Hope this helps.
<u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d"><u></u> <u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d">Mark</span><span style="font-family:"Calibri",sans-serif;color:#1f497d"><u></u><u></u></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1f497d"><u></u> <u></u></span></p>
<div>
<div style="border:none;border-top:solid #e1e1e1 1.0pt;padding:3.0pt 0cm 0cm 0cm">
<p class="MsoNormal"><b><span lang="EN-US" style="font-size:11.0pt;font-family:"Calibri",sans-serif">From:</span></b><span lang="EN-US" style="font-size:11.0pt;font-family:"Calibri",sans-serif"> <a href="mailto:adegenet-forum-bounces@lists.r-forge.r-project.org" target="_blank">adegenet-forum-bounces@lists.<wbr>r-forge.r-project.org</a> [mailto:<a href="mailto:adegenet-forum-bounces@lists.r-forge.r-project.org" target="_blank">adegenet-forum-<wbr>bounces@lists.r-forge.r-<wbr>project.org</a>]
<b>On Behalf Of </b>Nikki Vollmer<br>
<b>Sent:</b> 30 January 2018 18:08<br>
<b>To:</b> <a href="mailto:adegenet-forum@lists.r-forge.r-project.org" target="_blank">adegenet-forum@lists.r-forge.<wbr>r-project.org</a><br>
<b>Subject:</b> [adegenet-forum] How to interpret Density Plot for K=2<u></u><u></u></span></p>
</div>
</div><div><div class="h5">
<p class="MsoNormal"><u></u> <u></u></p>
<div id="m_-9004693243091395469divtagdefaultwrapper">
<p><span style="font-family:"Calibri",sans-serif;color:black">Hi,<u></u><u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black"><u></u> <u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black">I am trying to analyze ~200 RADseq loci for ~200 individuals. STRUCTURE results suggest the best number of populations given the data is 2. Pairwise Fst values are quite low for my taxa (<0.003)
with pvalue 0.01802. I was trying to do a DAPC on this same data to compare results. DAPC similarly suggested the best # of clusters is 2 and I was able to plot a 1-dimensional density plot for the one DF I kept (attached). However, I am not sure how to
interpret the plot. Is it correct to say that because the two peaks do not overlap that suggests the 2 clusters are quite differentiated from one another (similar to two clusters on a scatter plot being in opposite quadrants)? (...or is that logic flawed?)<u></u><u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black"><u></u> <u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black">I am trying to figure out if these 2 groups are very genetically differentiated or not, and I am not clear what the density plot is supporting/suggesting. <u></u><u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black"><u></u> <u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black">I very much appreciate any guidance on this matter!<u></u><u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black"><u></u> <u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black">Thank you,<u></u><u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black">Nikki<u></u><u></u></span></p>
<p><span style="font-family:"Calibri",sans-serif;color:black"><u></u> <u></u></span></p>
</div>
</div></div></div>
Inverness College UHI, a partner in the University of the Highlands and Islands <a href="http://www.inverness.uhi.ac.uk" target="_blank">www.inverness.uhi.ac.uk</a> Board of Management of Inverness College (known as Inverness College UHI), Scottish Charity No SC021197.
</div>
<br>______________________________<wbr>_________________<br>
adegenet-forum mailing list<br>
<a href="mailto:adegenet-forum@lists.r-forge.r-project.org">adegenet-forum@lists.r-forge.<wbr>r-project.org</a><br>
<a href="https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/adegenet-forum" rel="noreferrer" target="_blank">https://lists.r-forge.r-<wbr>project.org/cgi-bin/mailman/<wbr>listinfo/adegenet-forum</a><br></blockquote></div><br></div>