I am working genotyping-by-sequencing data of Zebrafish and it seems that there is no training/truth variant set available for applying GATK Variant Quality Score Recalibration (VQSR). I only found the dbSNP vcf files.
Has anyone performed VQSR on Zebrafish data and/or can suggest a good variant set for training/validation?
I am thinking of creating my own test datasets filtering variants using the evidence status (http://www.ensembl.org/info/genome/variation/prediction/variant_quality.html). Has anyone experience with this?