It was mentioned in a genetics class that in a genetic association analyses of a trait with all SNPs, it is possible to compute the expected effect of each SNP with the trait using the correlation structure between the SNPs and the size of the effect at the true causal variant.
Can you explain why this is so?
