EEMS-AROUND-THE-WORLD

Goal

This pipeline was built for the Peter et al 2019 manuscript on applying EEMS to a number of human populations and compares the results to PCA on the same datasets. The pipeline share here includes a workflow that comparisonn between several additional methods (listed below).

Reproducing results from Peter et al. 2019

As some of the data used requires permission, we are not free to redistribute it. To re-generate all figures from the paper, it will be necessary to

1. acquire access to all data and create the master data set as described in the merge-pipeline
2. change paths in config/config.json to reflect your working environment
3. run snakemake all

Implementation details

Genotypic data is stored in plink format. Metadata/location data is stored using the PopGenStructures data format, with some minor (recommended) changes. The pipeline is implemented using Snakemake, using python for most data wrangling and R for most plotting

Implemented methods

• EEMS
• flashpca
• admixture
• pong visualization of admixture
• TESS3
• treemix
• Spacemix
• conStruct
• FST (using plink)