Computer Simulation Linked to Gene Information Databases as a Strategic Research Tool to Evaluate Molecular Approaches for Genetic Improvement of Crops

Introduction Today, a wide range of molecular tools and powerful computers provides us with an unprecedented capacity to generate data that reveals the architecture of genomes, genes, traits, and how these contribute to the value of the germplasm units (populations, lines, varieties, hybrids, cultivars) used in breeding programs. Complementary work is characterising the key environmental variables that influence the expression and value of genes and improving our understanding of the composition and structure of the Target Populations of Environments (TPE) faced by breeding programs. Combining these two avenues of research provides a foundation for understanding gene function and expression and definition of the value of the alternative alleles of genes within the context of our reference breeding populations and target genotype-environment systems. The availability of these data, in appropriate information management systems, promises opportunities for accurate, rapid multi-gene manipulation strategies, with greater precision than previously achievable. The processes for generation of these gene-environment data are critical to the design and implementation of optimised molecular-conventional breeding strategies that build on the successes of our current suite of strategies. However, the data generation processes by themselves are insufficient to enable the implementation of the range of possible molecular breeding strategies. Mature breeding programs examining the potential of molecular tools to assist in achieving breeding objectives have proceeded beyond the phase of concentrating solely on data generation. There is recognition of the need to link the growing body of data on the value of alternative alleles with the germplasm units used within breeding programs. Without a link between the data generation processes and the units of manipulation in the breeding program the data are not accessible as real-time information to assist the breeder with cross and selection decisions. The large private sector breeding programs are well advanced compared to their public domain counterparts in this bioinformatics research and development area.