Here video of Christophe Lambert's excellent presentation: Dealing with Big Genomic Data: Towards Predictive & Preventative Medicine which is of interest to both the genomics folks, but also the architecture folks as well. It documents the accelerating changes happening in the life sciences, as well as our decelerating abilitty to deal with them.
Tip of the hat to Gerry Higgins for pointing out eRoom's law: Moore's law spelled backwards, showing our decreasing rate of acceptance of science in medicine, in addition to the inherent 17 year cultural lag in health care innovation for new knowledge generated by randomized controlled trails to be incorporated into practice. (See Crossing the Quality Chasm)
So, rather than just looking at genomics as a side-feature to the "real" EHR, we should use it as a case study on how to accelerate the adoption of science information into health care. Rather than fantasizing that we are going to achieve "semantic interoperability" based on boolean logic, we are going to have to deal directly with fuzziness and uncertainty that are impossible to explore through randomized clinical trials. We simply don't have the intellectual paraphernalia to deal with this level of complexity, but this is what we need for the future of health it...
good food for thought... and Thanks, Gerry for organizing this event.