Interview with Matt Might
pl-enthusiast.net
pl-enthusiast.net
Matt is a true mensch and a hero in rare disease diagnosis. What words I have are insufficient to express my deep admiration and appreciation. Well worth your time to read a bit about Matt's work.
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!
ps: it seems like another point for polymathy.
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.
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.
I'll keep your mail in mind too.
One of the problems with the precision medicine initiative is that medicine only loosely incorporates the actual findings of physiology. For example, old people routinely get their steroid hormone systems kneecapped with statin drugs. Various media reports in the last few years have revealed that there is very little benefit to suppressing cholesterol levels (women receive no benefit at all), but the drugging of old people continues.
1. Matt is one of two classmates that I remember from that high school. I switched schools for the last two years, didn't integrate well, and was recovering from an injury for half of my senior year. But Matt was friendly, and loaned me a book about C++, which I eventually mailed back to him.
The other person was a neighbor. His name was Chris, and he had a nicely-restored orange truck.
2. The problem is not with "precision medicine", but with medicine itself. The name of this project implies that they're going to turn medicine's hunting rifle into a sniper rifle. But conventional medical practice has inherited a bunch of assumptions that are basically wrong. Standard medicine's "rifle" is just a crummy old shotgun that's lucky to hit anything.
If you get shot in the head, go to the trauma center - they do good work, and can possibly save your life. If you have a chronic disease, get a dozen second opinions, including from non-medical doctors. If you have a psychiatric problem, avoid psychiatrists as if your life depends on it: conventional psychiatry is a disaster-in-progress.
The precision medicine effort is an opportunity to clean Medicine's house. The drug companies won't be happy when their businesses wither away, but maybe the lawyers will help them go quietly into the night.
39 page JFP version:
http://arxiv.org/pdf/1107.3539v1.pdf
12 pager from ICFP:
http://arxiv.org/pdf/1007.4446.pdf
8 page CACM highlight:
http://www.ccs.neu.edu/home/dvanhorn/pubs/vanhorn-might-cacm...