Fast multipoint linkage analysis and the program Allegro

¥ The first two authors have contributed equally to this work 2 Summary Several improvements to computational algorithms for multipoint linkage analysis are developed. The new algorithms achieve considerable speed-up over previous ones, and allow larger families to be analyzed. The algorithms have been implemented in a computer program, Allegro. Allegro has the same basic functionality as the well known Genehunter program, includes the features of Genehunter-Plus, and contains some added functionality. Among the supported features are parametric LOD scores, al-lele sharing LOD scores, NPL scores and haplotyping. The program is simple to use and accepts the same data file format as Genehunter. It has been used extensively and typical speed-up compared to Genehunter is 30-fold. The linkage analysis of Allegro and Genehunter involves three steps. Firstly, determination of single point probabilities of individual inheritance vectors, secondly multipoint calculation, where genotype information on neighboring markers is taken into account, and thirdly score calculation. The bulk of the time used by Genehunter involves steps 1 and 3 and it is here that the improvements are greatest; the key idea is to make use of tree traversal to avoid repeated calculation for similar inheritance vectors. In addition the fast Fourier transforms in step 2 have been made faster in the new algorithms. Finally, a new technique, founder couple reduction, halves the total execution time and memory requirement for many large pedigrees.