Skip to content

Latest commit

 

History

History
61 lines (47 loc) · 2.32 KB

TreeMix.md

File metadata and controls

61 lines (47 loc) · 2.32 KB
Raw

TreeMix Tutorial

By Erika Landa & Pavel Salazar-Fernandez (epsalazarf@gmail.com)

Human Population and Evolutionary Genomics Lab | LANGEBIO

About

Documentation: TreeMix

This document explains how to manipulate PLINK data for a TreeMix analysis and plotting.

Preparing the Data

Requires:

  • PLINK (version 1.07+)
  • Python 2.7+

  1. Get the PLINK binary files (.bed, .bim, .fam) to be analyzed. The dataset must contain only samples of interest.

  2. Generate a population tags file containing three columns: FAMID and ID (the same as the samples from the plink files), and an added column POP that tags each sample to a population. If the FAMID is the POP tag, you can generate thepopulation tags file like this: awk '{print $1,$2,$1}' DATA.bim > poptags.txt Example (without header): FAMID ID POP fam1 ind1 popA fam2 ind2 popA fam3 ind3 popA fam4 ind4 popB fam5 ind5 popB ... ... ...

  3. Run the following command to generate a cluster file (.clst) with PLINK: plink --bfile [DATA] --within poptags.txt --write-cluster --out clusterstep

  4. To generate an stratified frequency file, run: plink --bfile [DATA] --freq --missing --within clusterstep.clst --out [DATA]

  5. Compress the resulting file into a gzip format: gzip [DATA].frq.strat

  6. Run the given script plink2treemix.py to convert the dataset to a TreeMix input file: python plink2treemix.py [DATA].frq.strat.gz [DATA].tm.in.gz

Performing the Analysis

  1. Having the input file ready, a basic maximum-likelihood tree can be generated: treemix -i [DATA].tm.gz -o [DATA].tm.m0
  2. To account for migration events, a -m parameter followed by the number of migrations allowed can be added to the basic command. treemix -i [DATA].tm.gz -m 1 -o [DATA].tm.m1 treemix -i [DATA].tm.gz -m 2 -o [DATA].tm.m2 treemix -i [DATA].tm.gz -m 3 -o [DATA].tm.m3

Plotting the Results

  1. Load the given plotting_funcs2.R script inside an R environment. source("plotting_funcs2.R")
  2. Use the plot_tree function to generate the graphics: plot_tree("[DATA].tm.m0")

Auto-Mode Script

The following script performs all previous steps automatically. It requires that the FAMIDs are already replaced by POP tags. The max number of migrations can also be set (default: 3).

[Work in progress]