I want to use GATK to estimate cross-sample contamination for Whole Genome Sequencing data.

The specific tool is ContEst and it is run with:

 -jar GenomeAnalysisTK.jar \
 -T ContEst \
 -R reference.fasta \
 -I:eval tumor.bam \
 -I:genotype normal.bam \
 --popfile populationAlleleFrequencies.vcf \
 -L populationSites.interval_list
 [-L targets.interval_list] \
 -o output.txt

The --popfile option requires a vcf file representing population allele frequencies. Where can I get it? Is it the one I can get from dbSNP?

I'm using the human_g1k_v37.fasta reference, a version of hg19, "for the phase1 analysis we mapped to GRCh37. Our fasta file which can be found here called human_g1k_v37.fasta.gz, it contains the autosomes, X, Y and MT but no haplotype sequence or EBV".


1 Answer 1


On the GATK forum they've recommended the population stratified VCF file for this purpose.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.