What kind of study would be the best/cheapest way to identify the genetic risk alleles for a common disease, which you believe (from segregation analysis) to be polygenic in nature?