Establishing a valid approach for estimating familial risk of cancer explained by common genetic variants