How do I perform a plot from admixture analysis with multiple runs

Question

[alexis.diaz]

Hi all,

I am wondering how I can perform a plot from admixture analysis with multiple runs. I have obtained 100 runs for each K value from 1 to 10. So my question is what to do with the multiple .P and .Q files obtained from this analysis to make the plot.

Any suggestion?

Answer 1

[Aaron]

You can also check out the package pong (links to GitHub repo and original publication), which can process many hundreds of Q-matrices (orders of magnitude faster than CLUMPP/CLUMPAK) across multiple values of K simultaneously.

pong highlights any genuine differences between replicates (multimodality) while collapsing redundant information (e.g. ADMIXTURE results that are identical once you account for label-switching). It also aligns runs across K.

Answer 2

[alexis.diaz]

I found the answer, you need first to move all *Q files from each K into a single folder. Next, you should process this folder using 'pophelper' in R, using the following command:

remotes::install_github('royfrancis/pophelper')

library('pophelper')

setwd('/Users/alexisdiaz/Desktop/popgen/admixture_analysis/admixture_macosx/results2')

afiles <- list.files(path=".", full.names=T)

alist <- readQ(files=afiles)

readQ(files=afiles, filetype = 'basic')

clumppExport(alist, exportpath=".")

Pophelper will create 'paramfiles' and combined txt files in folders per each K. Next, make sure you place clumpp.exe file into each folder, click on the executable file and it will generate three other files. Use the *-combined-merged, copy values and plot it in excel.