X’Prime builds TGRM, a mate selection system that balances genetic gain against the cost of inbreeding, with room for the gene-marker data that’s reshaping what breeders actually decide on. Twenty-five years of commercial pedigree work, rebuilt from the ground up.
Give it a pedigree and whatever you know about your animals, EBVs, genomic breeding values, discrete gene markers, and TGRM works out which animals to mate with which. It lifts the traits you care about without quietly concentrating the gene pool, keeps carrier loci and allele targets under control, and shows you the trade-offs instead of hiding them. The maths are the ones the field has settled on, optimal contribution selection, Meuwissen-Luo inbreeding, Colleau relationship products, differential-evolution mate allocation, now running as TypeScript and a small Rust engine compiled to WebAssembly. No server required. No Fortran. No install.
Parse a pedigree file, build the DAG, compute inbreeding coefficients and candidate relationship submatrices. Ready to fold in GRM from SNP data where you have it. Fifty thousand animals in a few seconds.
Sweep the gain-versus-diversity frontier with a quadratic programme. Pick your target on the curve and see exactly what it costs you, in coancestry, in gain, in which sires get used.
Differential evolution solves the actual mating list. Per-sire limits, trait emphasis, progeny inbreeding avoidance, and first-class constraints for discrete loci, carrier avoidance, allele targets, genotype frequencies. QTL columns parse natively.
The client/server system you already use isn’t going anywhere. Your reports, datasets and consultant logins continue to work. The new browser version runs alongside it, try it on the same data, no migration required.
X’Prime was founded in 2001 by Rod Vagg and Susan Meszaros to bring TGRM, developed at the University of New England by Professor Brian Kinghorn, out of Fortran and into the hands of breeders. Twenty-five years on, we’ve served 34 breeding organisations across six species, processed two million candidates, and run seventy thousand optimisation reports. The field has moved on: breeders now routinely have SNP chips, gene markers, and genomic breeding values that the old academic tools weren’t really built around. We’re a small, stubborn company. We think mate selection software ought to be clean, fast, and honest about what it’s doing, and it ought to work with the data breeders actually have now, not what they had in 2001. So we rebuilt it.