Some thoughts triggered by the Barcelona #ECTB2018 meeting. 1) Genomic Medicine is here, *now*, certainly for rare disease. Exome or Genome (doesn't really matter which) is *cost effective* at diagnosis and informing management of rare diseases.
This means ~2% of live births in *every* country (certainly every country with a developed healthcare system) should have an exome or genome, plus parents (usually ~2.5 genomes per suspected rare disease genome)
This is solid, replicated, multi-system experience, and means that over time, ~5% of every population should have their genome sequenced. Every country should be actively planning this, and working out how to operationalise it inside their healthcare system.
2) "Beyond" rare disease, the low hanging fruit is informing decisions where clinicians are forced into a choice. So - for Prostate cancer, watchful waiting vs active treatment (great talk by Paul Boutros); for AML, transplant or chemo (@MoritzGerstung's work).
The "bar" here is very different from establishing a new treatment pathway. This also is true for the PRS move in common disease - stratifying MI risk for people in the 50s to go to statins, or stratifying Breast cancer screening programs is the near term goal
3) This is an evolution of medicine, not a revolution - genomics is an amazing tool, not a oracle (a common theme from me!). I often go back to the uptake of Imaging into Medicine (Here's a blog post in 2015: ewanbirney.com/2015/10/genomi…) which was a ~30 year bumpy ride
Imaging in Medicine is a useful (though not perfect) analogy to genomics. It is technical (in particular MRI sequences etc), has a medical speciality (Radiology) but is also routinely used by a broad set of clinicians, who know when and where to use it in their own practice
More broadly we need an *engineering* as well as a *research* mindset in this - indeed, the engineering mindset (operational, robust, tested) is more important in this phase than the research - though both are critical, like any part of complex technical delivery
5) If you thought we had a capacity problem in computational biology it is going to get *extreme* as every healthcare system has to get this skill set into medical practice (taking the radiology example, perhaps the same number as radiologists??)
This is a *huge* number of people, and we're just not teaching and training at with the right capacity. This leads to frazzled bioinformaticians, often without leadership/hard won experience, and messy situations. This is probably our most urgent problem!
