A simple python script to contaminate Eigenstrat genotypes.
A tool artificially contaminate the genotypes of multiple sample individuals with genotypes from a contaminant individual, at different rates of contamination.
Available options:
-h, --help show this help message and exit
-i <INPUT FILES PREFIX>, --Input <INPUT FILES PREFIX>
The desired input file prefix. Input files are assumed
to be '<INPUT PREFIX>.geno', '<INPUT PREFIX>.snp' and
'<INPUT PREFIX>.ind'.
-o <OUTPUT FILES PREFIX>, --Output <OUTPUT FILES PREFIX>
The desired output file prefix. Three output files are
created, '<OUTPUT FILES PREFIX>.geno', '<OUTPUT FILES
PREFIX>.snp' and '<OUTPUT FILES PREFIX>.ind'.
-s <SAMPLE1,SAMPLE2,SAMPLE3,...>, --Samples <SAMPLE1,SAMPLE2,SAMPLE3,...>
The sample individual, whose genotypes will be
contaminated.
-c <CONTAMINANT>, --Contaminant <CONTAMINANT>
The contaminant individual, which will be used to
contaminate the genotypes of each <SAMPLE> at the
specified rate(s).
-r <RATE1,RATE2,RATE3,...>, --rates <RATE1,RATE2,RATE3,...>
A comma separated list of contamination rates.
-n <nrReps>, --nrReps <nrReps>
An integer value specifying the number of replicate
contaminated genotypes to be created per contamination
rate [1].
-v, --overlapOnly When provided, only SNPs that are present in both
Contaminant and Sample will be contaminated, while all
other SNPs will be set to missing. This ensures
conaminant genotypes will not be diluted due to lack
of coverage overlap in the Sample and Contaminant.
The provided script will contaminate the genotypes of a given sample individual
in an Eigenstrat dataset at a specified rate, to match the genotypes of a
specified contaminant individual. Missing data in the sample individual will be
kept as missing. Heterozygotes in the contaminant will be pseudohaploidised. A
number of replicates can be specified for each contamination rate. Contaminated
individuals will be saved as a separate individual within the resulting
Eigenstrat database, following the naming scheme Sample_ContaminationRate_ReplicateNumber.
The population assigned to each replicate follows a similar naming scheme, but without a replicate number.
In cases where the contaminant individual has a missing genotype, the default behaviour is to add missing genotypes at the specified contamination rate. The -v/--overlapOnly option can be provided to override this default and set all genotypes that are missing in either the sample or contaminant to missing. In cases where both the sample and contaminant have incomplete coverage (as is often the case with ancient individuals) this will lower overall SNP coverage, but keep the resulting contamination rates consistent.