How well do we understand the functions encoded by our genome? Certainly, comprehensive functional informa tion about proteins, including the impact of mutations, is complete for relatively few genes. The development of high-throughput systems for biochemistry and enzymology could have a dramatic impact on this deficiency and would add vitality to these areas of scientific endeavor. Efforts that annotate regulatory protein binding sites, sites of RNA-mediated regulatory mechanisms, and other motifs that contribute to transcriptional regulation in the human genome must continue. Improved understanding of these regions, and thus their annotation, will require the power of model-organism-based systems to identify and characterize functional proteins or mechanisms that are shared with humans. We also must transfer these findings into human cell experimental systems that allow researchers to examine the impact of the mutations or other alterations of the genome on cellular pathways and the resulting disease biology. With functional consequences in hand, we will begin to understand and associate the clinical validity of genomic variants, effectively enabling the correlation of variant (s) with the resultant phenotype (s). If our efforts to improve the human reference sequence quality, variation, and annotation are successful, how do we avoid the pitfall of having cheap human genome resequencing but complex and expensive manual analysis to make clinical sense out of the data? One approach would emphasize the development of ‘clinical grade’inter pretational analysis pipelines to perform much of the initial discovery from datasets derived from massively parallel sequencing [8]. Although such pipelines already exist in the research setting [9], manual checks and orthogonal validation of variants are required because of the ongoing development of the analytical approaches. Towards patient diagnoses, such validation could initially be performed in a clinical laboratory medicine setting, but ultimately we must develop sophisticated analytical approaches and quality filters that enable high-confidence variant detection solely from the primary data. All discovered variants would then be interpreted in the context of the ever-improving human genome annotation and evaluated in the contexts of medical genetics, of demonstrated clinical validity, and of the pharmaceutical databases (when appropriate), to identify causative or therapeu tically actionable genes. Ultimately, as in medicine today, the results will require interpretation by a physician, which raises a separate but equally important issue: the significant need to develop and implement training programs in genomics for medical professionals. Pathologists and genetic counselors will be the first in line for training programs focused on genomic diagnostics, and improving the genomics education of medical students will also be a first priority. More challenging will be the genomics education of practicing physicians and other medical professionals, many of whom do not require genetics to perform their valuable role in health care daily, but who will be confronted in the near term by increasingly well informed patients who expect their doctors to be as well versed as they are about genome-guided diagnosis and treatment. A final word on the important topic of patient access to genome-guided medicine seems necessary and appropriate. The current high cost of whole-genome sequencing and analysis relative to most clinical diagnostic assays, coupled with the fact that these costs are not currently reimbursed by insurers, might mean that only those with the means to pay for the test will be allowed access …