There's a lot of really basic modelling of genes that can be done for rare Mendelian diseases using formal language theory which is surprisingly not there yet.
I am working closely with some Genomics England people (100k genomes, many from rare disease patients), and as someone with a pure CS background but also some years of biology the rather primitive state of the art never ceases to amaze me.
Matt mentions pi-calculus, but much more basic things are not there yet!
Nope, but we will have a Nature paper out quite soon.
If you want some info, ping me. Contact email is on my profile.
I think we desperately need some startups in this arena that lean towards open knowledge. I know some very well funded ones in stealth mode which are producing pretty awful and closed prediction models. I don't want a future where my doctor uses information coming from a blackbox.
The point about polymaths is good indeed. But they are hard to find. I have some funding, but found it incredibly hard to hire / team up with someone good enough. Most get lured into finance, which pays well but might not be the most useful thing for humankind.
Polymath at the individual or company level (cross domain would be more appropriate ? DNA Crystallography comes to mind)
I couldn't agree more with you about openness of advances in medicine, it saddens me that too few people care about this. I wish I was in a position to help in any way.
Both. Well, I think focussed individuals with 1 area of expertise are fine. But you need someone who is at least a a very good computer scientist, has some formal background in probability theory and ideally grasps genetics quite well (which is far from trivial as it's a moving target).
I am working closely with some Genomics England people (100k genomes, many from rare disease patients), and as someone with a pure CS background but also some years of biology the rather primitive state of the art never ceases to amaze me.
Matt mentions pi-calculus, but much more basic things are not there yet!