Sunday, September 15, 2013

Things I learned this week at the Packard meeting

For the 25th anniversary of the amazingly awesome David and Lucille Packard Foundation fellowships, I was fortunate (and as always very grateful) for the opportunity to go to their annual meeting and listen to talks from incoming and outgoing Fellows.  These meetings are tremendous - a very rare chance to hear 20 minute talks on topics across science, engineering, and math, aimed at the technically literate non-expert.  Back in the dim past of this blog, I've posted about these before (here, here, and here).  Here are some take-away facts I learned this time around:
  • By very narrow targeting of specific pathogens, it might be possible to remove some of the evolutionary pressure (exerted by horizontal gene transfer [something I'd never learned about] from your gut bacteria) that leads to antibiotic resistant strains.
  • It's possible to use ideas from superresolution microscopy and principal component analysis to improve structure determination in materials characterization.
  • Using small molecule dyes, it is possible to use optical processes to turn the tables on some chemical reactions, favoring "anti-Markovnikov" selection, rather than Markovnikov rules (where reaction sites are determined by permanent dipole moments of bonds).
  • Sometimes cells can recognize themselves (and distinguish between themselves and close relatives) using proteins based only on one or two genes.
  • I'm used to thinking about coupling two (identical) resonators and getting an energy splitting (like bonding/antibonding orbitals).  I hadn't realized that using an effectively imaginary coupling means you can get a lifetime splitting (one long-lived, one short-lived mode).
  • You can tie vortex rings in knots.  Watch the videos!
  • Greenland has not been ice-free for at least 350,000 years, and radioactive dating based on dust captured in the ice makes it possible to untangle even faulted or folded ice cores.
  • Monsoons are complicated, even if you model a completely water-covered idealized planet.
  • Every time a pair of neutron stars collide, they produce about one Jupiter mass worth of Au, while a core-collapse supernova makes about one lunar mass worth of Au.  As a result, even though colliding neutron stars are rare, half of the gold out there came from them.  (In case you were wondering, in all of human history we have mined about 165,000 tons of Au.)
I've left out many others.  As always, very cool.

5 comments:

Anonymous said...

who gave the first talk, on the gut stuff?

Douglas Natelson said...

Anon., it was Doug Mitchell at UIUC.

Anonymous said...

"I'm used to thinking about coupling two (identical) resonators and getting an energy splitting (like bonding/antibonding orbitals). I hadn't realized that using an effectively imaginary coupling means you can get a lifetime splitting (one long-lived, one short-lived mode)."

Can you point me to where I can learn more about this or tell me who gave the talk?

Thanks

Douglas Natelson said...

Anon@9:49, I struck out looking for a reference, but it was Milos Popovic at Colorado who gave the talk.

Anonymous said...

It seems similar to the coupling of molecules true transition dipole moments, i.e. j-aggregates. As consequence you have to “new” electronic states (normally, only one transition is allowed) and some nice QM effects as delocalized hole-electron pair and the quenching of the vibronic signature.