Does someone know of a software that can simulate genotypes for a variant with AC and AN counts, given a set of variants? I am assuming AC will help generate the heterozygous variants [0/1] or homozygous [1/1] and AN can give the number of samples needed in a VCF format.
2 answers
There are more direct ways to do this, but
plink --simulate-qt [parameter file] acgt --simulate-n [number of samples] --recode vcf --real-ref-alleles --out [output file prefix]
where the parameter file has a single line of the following form:
[number of variants] [variant name prefix; if this is "var", your variants will be named "var_0", "var_1", ...] [minimum alt allele fraction] [maximum alt allele fraction; set this identical to the minimum if you want all the MAFs to be the same] 0 0
should work. (See https://www.cog-genomics.org/plink2/input#simulate for more details)
RTG Core contains a command rtg samplesim which takes as input a VCF containing AF INFO annotations, and outputs a VCF that has a new sample column added, with each genotype randomly drawn according to the allele frequency. If you can get your AC/AN counts converted to AF allele frequency then you're done.
Log in to answer this question.
Doesn't GATK provide those field? I think it does: https://www.broadinstitute.org/gatk/guide/article?id=1268
I am trying to find a simulator software not a variant caller given bams [unlike a GGA call].