A simulation approach for analyzing genomic data using a package of specific FORTRAN90 functions

A panel of FORTRAN 90 functions was developed to simulate the distribution of bi-allelic (e.g., SNP) genetic markers along a defined genome and the distribution of their alleles in a given population. The simulation program used 3 parameters, those related to the species studied (number of autosomes, average length of autosomes, average number of crossovers by chromosome), the number of markers and those related to the studied population (pedigree). The simulation proceeds in 3 steps: a) random choice of marker positions and allelic frequencies for the minor allele of each marker (range: 0.05 to 0.475), b) simulation of
genotypes of the ancestors in the pedigree based on randomly chosen allelic frequencies and c) planned mating of the ancestors according to the pedigree and according to the average crossover rate as a genetic
recombination parameter. The simulation returns a fully-genotyped population. This method is flexible because it can be applied to a wide range of cases (not restricted to a single species) and the FORTRAN
functions can be extended and used to simulate phenotypes. It is also realistic, because it performs mating plans and selection of animals based on real pedigrees. Development of this simulation panel was the
first step in research around advanced methods to compute and invert genomic relationship matrices.