Given vectors of family sizes and number of affecteds per family, calculates the risk of disease recurrence in an unaffected person, conditional on a family having at least k affected members. Methods also model heterogeneity of disease risk across families by fitting a mixture model, allowing for high and low risk families.