Comments on: Chemistry of the future: 3D models and augmented reality http://boingboing.net/2011/09/26/chemistry-of-the-future-3d-models-and-augmented-reality.html Brain candy for Happy Mutants Sun, 19 May 2013 03:01:00 +0000 hourly 1 http://wordpress.org/?v=3.4.1 By: Morgan Hocking http://boingboing.net/2011/09/26/chemistry-of-the-future-3d-models-and-augmented-reality.html#comment-1226213 Morgan Hocking Tue, 27 Sep 2011 04:38:00 +0000 http://boingboing.net/?p=120075#comment-1226213 am i the only one that thought this was an augmented reality ball of gummy bears? am i the only one that thought this was an augmented reality ball of gummy bears?

]]> By: Alan Chen http://boingboing.net/2011/09/26/chemistry-of-the-future-3d-models-and-augmented-reality.html#comment-1225898 Alan Chen Mon, 26 Sep 2011 22:39:00 +0000 http://boingboing.net/?p=120075#comment-1225898 It's a little bit of both.  The search space for docking of two rigid bodies is unexplorably huge.  How huge?  http://arstechnica.com/business/news/2011/09/30000-core-cluster-built-on-amazon-ec2-cloud.ars  This huge ( I'm 90% sure this was a docking calculations).    The systematic optimization algorithms work like dumb jigsaw puzzle players.  Choose a random orientation, press to fit, if no, rotate slightly, try again.  Move it slightly, try again.  If you could allow a human to direct the process however (even just by giving a good initial guess), then the search has a much better chance of succeeding.  This interface would be a reasonable way to do that. The  algorithms rapidly become worse and worse the more realistic you make this problem. If you let one molecule be slightly flexible, as they are in real life, human intuition can easily outperform the existing algorithms. (see recent press surrounding FoldIt) . http://www.nature.com/nsmb/journal/vaop/ncurrent/full/nsmb.2119.html It’s a little bit of both.  The search space for docking of two rigid bodies is unexplorably huge.  How huge?  http://arstechnica.com/business/news/2011/09/30000-core-cluster-built-on-amazon-ec2-cloud.ars  This huge ( I’m 90% sure this was a docking calculations). 
 
The systematic optimization algorithms work like dumb jigsaw puzzle players.  Choose a random orientation, press to fit, if no, rotate slightly, try again.  Move it slightly, try again.  If you could allow a human to direct the process however (even just by giving a good initial guess), then the search has a much better chance of succeeding.  This interface would be a reasonable way to do that.

The  algorithms rapidly become worse and worse the more realistic you make this problem. If you let one molecule be slightly flexible, as they are in real life, human intuition can easily outperform the existing algorithms. (see recent press surrounding FoldIt) . http://www.nature.com/nsmb/journal/vaop/ncurrent/full/nsmb.2119.html

]]> By: Restless http://boingboing.net/2011/09/26/chemistry-of-the-future-3d-models-and-augmented-reality.html#comment-1225895 Restless Mon, 26 Sep 2011 22:37:00 +0000 http://boingboing.net/?p=120075#comment-1225895 I don't know why, but this argument is giving me flashbacks to the program meant to figure out how to get the sofa up the stairs in <em>Dirk Gently's Holistic Detective Agency</em>. I don’t know why, but this argument is giving me flashbacks to the program meant to figure out how to get the sofa up the stairs in Dirk Gently’s Holistic Detective Agency.

]]> By: arthurolson http://boingboing.net/2011/09/26/chemistry-of-the-future-3d-models-and-augmented-reality.html#comment-1225792 arthurolson Mon, 26 Sep 2011 21:30:00 +0000 http://boingboing.net/?p=120075#comment-1225792  Of course you are correct.  The same program that we use to compute the interaction energy of the manipulated tangible models is used for automated computational docking.  Our open source AutoDock code has been downloaded by over 30,000 laboratories world-wide and is used for "virtual screening of very large chemical libraries (see for example our FightAIDS@Home project).  The purpose of the tangible interactions is quite different.  The tangible models have perceptual advantages over computer generated imagery, and allow "playful" exploration for quick "what if" type of questions.  We are exploring the augmented reality tangibles as novel computer interfaces both for research applications and for educational applications.  Of course you are correct.  The same program that we use to compute the
interaction energy of the manipulated tangible models is used for
automated computational docking.  Our open source AutoDock code has been
downloaded by over 30,000 laboratories world-wide and is used for
“virtual screening of very large chemical libraries (see for example our
FightAIDS@Home project).  The purpose of the tangible interactions is
quite different.  The tangible models have perceptual advantages over
computer generated imagery, and allow “playful” exploration for quick
“what if” type of questions.  We are exploring the augmented reality
tangibles as novel computer interfaces both for research applications
and for educational applications.

]]> By: public bizmail http://boingboing.net/2011/09/26/chemistry-of-the-future-3d-models-and-augmented-reality.html#comment-1225478 public bizmail Mon, 26 Sep 2011 19:12:00 +0000 http://boingboing.net/?p=120075#comment-1225478 Why is this a good idea? Or is it just a good promo for fund raising? If you have enough data to  perform the 3-D  printing,  you have enough data to render  the molecules in silico. Once  rendered in In silico, the computer can algorithmically  do a much more exhaustive search than a human, and without the stubby fingers getting in the way.  Nor is the analysis confined to the resolution of a web cam. Does this lab have any successes with this method that outperformed the obvious virtual approach? Why is this a good idea? Or is it just a good promo for fund raising?

If you have enough data to  perform the 3-D  printing,  you have enough data to render  the molecules in silico. Once  rendered in In silico, the computer can algorithmically  do a much more exhaustive search than a human, and without the stubby fingers getting in the way.  Nor is the analysis confined to the resolution of a web cam.

Does this lab have any successes with this method that outperformed the obvious virtual approach?

]]> By: awjt http://boingboing.net/2011/09/26/chemistry-of-the-future-3d-models-and-augmented-reality.html#comment-1225429 awjt Mon, 26 Sep 2011 18:50:00 +0000 http://boingboing.net/?p=120075#comment-1225429 The difference is that now, the energy calculations are much more precise.  And the chicks.  Don't forget about the chicks. The difference is that now, the energy calculations are much more precise.  And the chicks.  Don’t forget about the chicks.

]]> By: Art http://boingboing.net/2011/09/26/chemistry-of-the-future-3d-models-and-augmented-reality.html#comment-1225295 Art Mon, 26 Sep 2011 17:21:00 +0000 http://boingboing.net/?p=120075#comment-1225295 "3D models of augmented reality"-Starring Al Pacino.  HOO-WAAA! “3D models of augmented reality”-Starring Al Pacino.  HOO-WAAA!

]]> By: herrnichte http://boingboing.net/2011/09/26/chemistry-of-the-future-3d-models-and-augmented-reality.html#comment-1225215 herrnichte Mon, 26 Sep 2011 16:34:00 +0000 http://boingboing.net/?p=120075#comment-1225215 That's very convenient and impressive - however it's not a dramatically different approach to what we were doing with CPK models on the desk in front of (vector!) displays back in the stone age (c1988) That’s very convenient and impressive – however it’s not a dramatically different approach to what we were doing with CPK models on the desk in front of (vector!) displays back in the stone age (c1988)

]]>