Does the GENOtation web-based PCA tool work?

Question

[Getting there]

http://genotation.stanford.edu/#PCA

I was trying to get a nice PCA visual of some of samples on top of Hapmap data using the web-based Genotation tool from Stanford linked above. I converted my VCF into the 23andMe format like they asked, but no matter what samples I seem to enter into the tool, they are always cluster in the same exact spot, even if I know for a fact they are from different ethnic backgrounds, not related, and should be widely separate on in PCA1 vs PCA2 in the Hapmap data.

I can actually upload any random text file, it doesn't even have to contain the rsIDs and it will still cluster a sample in the same exact spot, the tool doesn't throw out any error messages, it just takes any file and plots it as if it did PCA analysis on it.

Does anyone know anything more about this tool, if it's even functional? Or better yet, have some alternative method I can use to do ethnicity PCA analysis?

Answer 1

[Kevin Blighe]

I have never used that service from Stanford. Please contact them directly for queries.

As an alternative, you can try to follow my tutorial: Produce PCA bi-plot for 1000 Genomes Phase III - Version 2

If you follow my tutorial, then you will have to merge your dataset to that of the 1000 Genomes at some point.

Kevin

[Getting there]

Thank you Kevin I will try your workflow