Improved allele frequency filtering and incomplete penetrance

[damiansm]

We could consider using the approaches described in https://www.nature.com/articles/gim201726 and exposed in the tool at http://cardiodb.org/allelefrequencyapp/ to calculated an expected MAF cutoff based on mode of inheritance, penetrance, disease prevalence and heterogeneity. Prevalence levels are available from Orphanet now or we could just a let a user enter the values and set MAF filters appropriately.