Taxonomizing Diseases: Exploring Genomic Nosologies Using Translational Bioinformatics
Broad classifications of disease, or nosology, have been proposed for over 300 years based on symptoms and anatomy, and have lost much relevance to basic science. We hypothesize that sufficient data from molecular measurements now exists to begin a broad reclassification of diseases. I will describe a general nosology built from a collection of nearly 2,000 publicly available gene-expression microarrays covering 75 afflictions across all specialties of medicine. While the state-of-the-art disease classification SNOMED-CT significantly resembles our nosology, validating the taxonomical work performed by pathologists for 77 years, I will show that this classification of diseases is superior because it enables identification of new disease biomarkers and genes that suggest new indications for existing pharmaceuticals, a novel form of pharmacogenomics. Our nosology confirms the ubiquitous role of inflammation in disease.
With NIH budget doubling and completion of the Human Genome Project, there is a need to translate genome-era discoveries into clinical utility. The difficulties in making bench-to-bedside translations have been described: comprehensive molecular studies on patients are expensive, and hospitals are not phenotypers. The nascent field of translational bioinformatics may help. Our lab has successfully built novel "clinical microarrays" from quantitative electronic medical record data on 5,200 patients over 7 years. We have found that these clinical microarrays can also be used to classify diseases into a nosology, and the nosologies built from clinical measurements and genomic measurements can be compared across diseases for the first time.
Over 120 years ago, John Gouley noted that "the nomenclature of diseases can be much improved only after a very great change for the better shall be made in the fundamental science and associated arts of medicine". Despite their limitations, it is remarkable that in the ten years since their creation, microarrays have already been used to study so many human diseases. It is only because data from these experiments are publicly available that we can now start to comprehensively reconsider the nature of disease itself.