In the early 90’s, as the first human genome project was initiated, an extensive effort has been placed on mapping the role of genes in the onset of disease. The genomic variants alone are not able to account for the changing of risk factors across the life span. Thus, the key goals to understanding human health and disease depend on the ability to access the ‘genotype-phenotype’ correlogram through multi-omics platform and the success of translating technological innovations (e.g., machine learning). We demonstrated that using two-dimensional T1-T2 correlational spectroscopy on a single drop of blood, highly time– and patient–specific ´molecular fingerprint´ can be obtained using home-built NMR-based micro scanner. Machine learning techniques were introduced to transform the correlational map into user friendly information for disease diagnostic and monitoring. The clinical utilities of this technique were demonstrated through the direct analysis of human whole blood in various physiological and pathological states (e.g., blood oxidation, diabetes mellitus, malaria).