A computerised method for calculating the probability of pedigrees from genetic data

Identification problems resulting from paternity disputes, criminal cases or disasters may be difficult to resolve on the basis of DNA if data from close relatives is missing or mutations need to be taken into account. Any one pedigree of interest may be consistent with a large number of combinations of genotypes for the missing data. The paper describes a computer algorithm designed to compute likelihood ratios intended to assist in identification cases. A real case serving as the main example is beyond the reach of simple 'tour-de-force' algorithms. The authors' approach is largely based on ordering computations efficiently as well as splitting pedigrees, enabling handling of complex problems.