Thread by @ewanbirney

The @uk_biobank is ... amazing and is basically, in my view, rebooting the science of human physiology.

Why? First it is just a really well phenotyped cohort at scale. Back in the early 2000s a number of people did key power calculations and amazing (or not so amazingly for epidemiologists) 500,000 was the minimum prospective cohort to have to impact common disease

Secondly the phenotyping has been done centrally and consistently, and some key imaging phenotypes have been done at scale. This is remarkable logistics, fund raising/arguments and delivery. *so* many things could have gone wrong which didn't.

Thirdly the health record linkage by the NHS has its limitations (I can see Cathy Sudlow's eye roll at this description) but it does work. (not as good as Denmark's EHR, but... not bad) Hospital Episode Statistics - diagnosis of disease in ICD world is actually pretty useful

The long haul of primary care data has finally landed in UK BioBank and this will take this to the next level. Denmark is always gone to have that longitudinal reach that no other country can match, but ... i've seen far worse ICD code/EHR link up scenarios.

Fourthly the size and length of time of UK BioBank means companies are pitching new ways to look at blood (in particular) and by the rules of UK BioBank this becomes open to other researchers after a specific time.

Fifthly UK BioBank is unfussy about the science on data access. Very straightforwardly you don't have to be mates with the lead investigators to gain access; you do have propose sensible science and you do have follow the rules (eg, disclosure etc).

Sixth - There is extensive genotyping, now exome, will be genome. And probably the interesting thing is that this is just one of 6 great things about UK BioBank.

Detractors will bring up ascertainment. Which - is fair - it can't be all things and to get the logistics and scale they had to go to a broad volunteer basis. But here is where you need to blend UKBioBank results with well ascertained cohorts (my favourite: @CO90s - "ALSPAC")

I have made great clinical colleagues - now friends - around UK BioBank - epidemiologists (so much epi. I love epi.) and then Cardiology and Opthamology I have gone deep(er) into.

I sometimes joke that if I could meet an enthusiastic clinician scientist into Toenail physiology and disease I would happily do toenail genetics to inform toenail physiology and disease...

Just being able to really nail down that this bit of physiology works like this... and impacts this disease A and B but not disease C and D... it really feels like ones just rolling back the clouds of ignorance.

Many many things went right with UK BioBank - hats off to the funders @wellcometrust and @The_MRC - to all the people (Tim Peake and Mark Effingham) who made it happen but, for leadership, science and vision - Rory Collins deserves so much praise.