Presenter: Ingrid Tarr (Victor Chang Cardiac Research Institute) Abstract: Whole genome sequencing has transformed our ability to detect associations between phenotypes and genetic variants, however, the amount of erroneous variant calls has also drastically increased. Even with low error rates, a significant quantity of called variants will be false positives. These are particularly concerning in unbiased genome-wide rare variant analyses, where a smaller number of false variants can have a meaningful impact on results while broad confirmation of variants is unfeasible. Currently, evidence informing rare variant filtering is lacking and there is no consensus regarding indicators of poor variants. The ability of common GATK metrics to discriminate true-positive and false-positive rare variant calls from samples sequenced in duplicate will be discussed.