Admixture Analysis

This method performs an admixture analysis using ngsAdmix, please see their tutorial page for full details on this method. This includes additional flags for filtering, stopping criteria, etc.

---

Note:

This method requires output from the Genotypes function to run

Please calculate genotype likelihoods in beagle format (by setting DO_GLF=2 in the Genotypes configuration file) before performing admixture analysis

---

Basic Usage

To run this method, use the following command

angsd-wrapper Admixture Admixture_Config

where Admixture_Config is the full path to the configuration file for the 2D site frequency spectrum and F_ST estimations.

Input files

All inputs should be specified in Admixture_Config.

Common Variables

This method does make use of Common_Config, those that are used are listed below:

Variable	Function
PROJECT	Name given to all outputs in ANGSD-wrapper
SCRATCH	Place to store files, the full path is SCRATCH/PROJECT/Admixture
N_CORES	Number of cores to use, please do not set above the limits of your system

Method-Specific Variable

This variable is specific to this method:

Variable	Function
LIKELIHOOD	The likelihood file calculated from the site frequency spectrum, this should be the <PROJECT>_SFSOut.beagle.gz file. Note that the

Method Parameters

The parameters for this method can be tweaked as necessary, they have been set for optimal generalized function:

Parameter	Function
K	Maximum number of ancestral populations to be inferred
MIN_MAF	Minimum minor allele frequency filter

Output files

To be fixed

Naming Scheme	Contents
PROJECT.k.filter	stuff
PROJECT.fopt.gz	Compressed estimate of frequency at each site for population of size k
PROJECT.k.log	Information about running of analysis for k populations
PROJECT.k.qopt	Admixture proportions for all individuals from k populations

Where k is the number of populations being tested.

Visualization

PROJECT.k.qopt files can be visualized with the Shiny graphing interface. A web browser with a graphical user interface is required.

Advanced Arguments

As mentioned at the top of the page, NGSAdmix has additional options for running the analysis, notably filtering criteria such as minLrt for minimum likelihood ratio values and minInd for removing sites having with a limited number of individuals with data. These can be easily added within the provided ADVANCED_ARGS='' line, using traditional command-line format. For example: ADVANCED_ARGS='-seed 100 -P 10 -tol 0.00005 -minLrt 0.1'