Saturday, February 25, 2012

Munching on Medical Data

ISEPP took the risk of inviting an industrialist versus an academician to address the Portland lay public, an audience reminiscent of Wittgenstein's Vienna by some accounts.  Speakers like to unveil stuff in Portland, as do movie people.  Several in our circle had known David Cox from their shared high school (same school as Tara's, and formerly a business school before joining the public fleet, accepting that charter).

Pfizer, and one extrapolates to other pharmaceutical companies, is seeing vast wastelands of useless patents in fields were the name of the game is to learn science, meaning one needs a culture of openness and open data (a true society).  You can try to play lip service to those ideals, and get away with it for awhile, but then the physics of the inevitable kicks in and your realize you're stalled and/or slipping backwards, forgetting your science, your know-how.  That sense of an impending dark age provokes an avoidance reflex.

Fortunately, we're surrounded by evidence of great civilizations, including some very recent ones, such as the Apollo Project, which put humans on the moon.  More to come (in the pipeline and/or incubator right now).

The giant project to decipher the genome has yielded big data for sure, but the statistical correlations with diseases are relatively slight, giving subtle clues as to mechanisms.  What you need to find are concentrations of shared diseases and find correlations there.

The individuals should be enrolled as responsible medical team members, to an MD level in specialty i.e. with respect to their own condition (MDs licensed to practice on people besides themselves are not permitted such narrow self-centeredness).

Dick Cox encouraged us to begin the ascent to a more enlightened society, not out of some altruistic do gooder sense so much as a need to get on with the business of learning science and really finding drugs that work.  We just don't have that many yet and the side effects are killing us.  Drug company people take drugs too, have an investment in being smart about self-administration.  Some top researchers have first hand experience with the diseases they treat.

DNA research has opened plenty of doors, but to walk into this new world of medical science we have to be free and open with medical records in a way that doesn't miscast either government, industry or insurance as ipso facto "the bad guy" i.e. there are legitimate get-the-job done configurations that aren't interested in wasting anyone's time on scandalously poor ethics or impossibly wasteful and/or ineffective treatments.

There's too much ignorance about the body and how it works, about nature, to squander precious lives on soap operatics and melodrama.  Dialing back to preventative therapies is still the best approach, but that means seeing the tendencies, knowing the signs.  When you're preventing, you're preventing against, which means the diseases have shown themselves.  This is more of an Asian attitude perhaps, in that you're not looking for "cures" so much as negotiated settlements, tolerable equilibria, steady states.

Happiness, including on into old age, is legit to pursue, and not just because the US Constitution says so.

At the dinner afterward, I asked if I were one of the bad guys per his model, in that I was in the business, at least until recently, of purging case records of identifiable traits, i.e. I was into anonymizing and pseudonymizing major data sets.  Yes, I was a bad guy.  He used the word "encrypt" but I don't see myself doing that so much as rendering petabytes of case data innocuous as a public good, e.g. not subject to blackmailing purposes or leading to divorce action because of revealed details in peoples private lives.

However, we've detoured into a technical footnote as the allegiance between medical sciences and free open source computing has only just begun, yet has been one of those exponentially more relevant curves.  Many people have no problem being up front with their identity when presenting with various diseases and will not mind having their DNA fingerprint a part of the public record.

Those less interested in helping science in that way will not be coerced into doing so as protecting record privacy and security also relate to integrity i.e. if the data gathering models are not themselves clear and public, then the outcomes data will not be trusted -- could be just someone's faked / randomized noise, for sale at high prices.

The ability to authenticate to trusted sources needs to be part of the model as well.  But then banking is not a complete stranger to these issues.  Bank your genome.  Consider it yours (whom else would it belong to -- nature herself maybe as you're her baby) and perhaps some of your income might come under a kind of "thanks for the memories" in the chart of accounts (i.e. thanks for the samples, the records).  Pay people for being who they are, what a concept.  That's usually more applied to horses these days.

I'm expecting some of the upcoming conference events on my calendar to give me more of a sense of Big Data and medicine.  I'd say Pfizer is definitely not alone in thinking Portland has a role to play in designing the data collection centers and sharing the data.  The local buzz matches the global in this case.   Something like Google Earth but for the body, built with filters so you might see different demographic cross-sections... such public goods with pro versions might be one outcome of the upcoming data crunching and visualizing era.

People with exotic diseases have much to teach us and should not have to worry where their next meal is coming from as medical science works with its students -- as long as they don't mind cafeteria food (we hope there's a Food Not Bombs like end of the spectrum, fresh and vegan, with some of the healthier patients doing their own prep and serving).  The Global U will have better and more humane infrastructure as we shluff off the old snake skin.  To be continued.

No comments: