A reorganization of genomic data into a simpler standard form leads to more transparent data analyses. The customary selection practice that focuses on high odds ratios loci is shown to be biased, reflecting quality of presently reported risk loci for T2D. A selection criterion, based on Shannon information theory, brings clarity to this issue and provides a rational and optimal basis for selecting potential risk loci. This is used to determine an optimal disease classifier. Within the framework of the FUSION database this leads to a relatively successful degree of T2D prediction and nearly an order of magnitude more effective in detecting T2D. Chromosome 7 is strongly associated with T2D. A hypothesis of this study is that the genomic disease signal is possibly weak, and instead of focusing on individual loci a collection of loci contribute to a composite Score, which functions as the determinant of disease or its absence.