Hans Rosling and the Joy of Stats

In 2010, the BBC produced a highly entertaining and informative documentary – The Joy of Stats (link here) with renowned world health expert and infectiously enthusiastic Hans Rosling.  Rosling is the founder of Gapminder, a nonprofit organization that provides amazing tools for visualizing data among other things.  Rosling’s fantastic TED talk (see here), which uses some of the tools, is a must see.  It will change the way you look at the world.  This link points to a panel discussion at the Lindau Nobel Laureate Meeting 2011.  A transcript can be found here.  In it, Rosling shows convincing evidence that the world population will peak out at about 9 billion because we have reached “peak child” – the number of children in the world below 15 is 2 billion and it is not growing.  He argues that women the world over are choosing to have fewer children.  The developing world is also getting healthier but they can’t afford all treatments; what the world needs is a new way to develop cheaper medical treatments.

Aspen summer program

The Aspen Center is a famous place where physicists congregate during the summer.  This summer they will be hosting a program on the physics of behavior (see the invitation letter from the organizers below).  This will combine neuroscience and animal behavior and should be very interesting.  I plan to be attending.  I encourage everyone to apply.

The Aspen Center for Physics will be hosting a 3-week long workshop on Physics of Behavior between May 27 and June 16, 2012, with an application deadline of January 31, 2012. The idea of the workshop stems from the understanding that the role of physics in biology is broad, as physical constraints define the strategies and the biological machinery that living systems use to shape their behavior in the dynamic, noisy, and resource-limited physical world. To date, such holistic, physics-driven picture of behavior has been achieved, arguably, only for bacterial chemotaxis. Can a similar understanding emerge for other, more complex living systems?To begin answering this, we would like to use the Aspen Center workshop to bring together a diverse group of scientists, from field biologists to theoretical physicists, broadly interested in animal behavior. We would like to broaden the horizons of physicists by inviting experts who quantify behavior of a wide range of model organisms, from molecular circuits to mammals. We would like to explore behavior as possibly optimal responses given the physical and the statistical structure of environment. Our topics will include, in particular, navigation and foraging, active sensing, locomotion and rhythmic behavior, and learning, memory, and adaptive behaviors.

As workshop organizers, we encourage you to apply.  We would also like you to encourage other people who are active in this field to apply.  We do need to be clear, however, that we cannot guarantee admission to the workshop.  Admission to the workshop is granted not by the workshop organizers, but by the Admissions Committee of the Center (with some input from the workshop organizers).  The Admissions Committee will endeavor to accommodate as many applicants to the Workshop as possible, but because of the constraints imposed by the rest of the AspenCenter for Physics program, they may not be able to admit everyone who applies.

We encourage you to visit the web site of the workshop here , and of the Center, http://www.aspenphys.org/, for more information and for application instructions. For those of you unfamiliar with the Center, it is located in lively and beautiful Aspen, CO. It’s a great place to work, to enjoy the mountains, and to bring  family. The Center partially subsidizes lodging for admitted participants. The Center requires that theorists commit for a minimum stay of two weeks, and a three week stay is preferred. Shorter durations are possible for experimentalists.

We hope you will choose to apply. Please don’t hesitate to contact us if you have questions.


The Organizers
Ila Fiete, UT Austin
Ilya Nemenman, Emory U
Leslie Osborne, U Chicago
William Ryu, U Toronto
Greg Stephens, Princeton U

New paper on steroid-mediated gene induction

A follow-up  to our PNAS paper on a new theory of steroid-mediated gene induction is now available on PLoS One here.  The title and abstract is below.  In the first paper, we proposed a general mathematical framework to compute how much protein will be produced from a steroid-mediated gene.  It had been noted in the past that the dose response curve of product given steroid amount follows a Michaelis-Menten curve or first order Hill function (e.g. Product = Amax [S]/(EC50+[S], where [S] is the added steroid concentration)..  In our previous work, we exploited this fact and showed that a complete closed form expression for the dose response curve could be written down for an arbitrary number of linked reactions.  The formula also indicates how added cofactors could increase or decrease the Amax or EC50.  What we do in this paper is to show how this expression can be used to predict the mechanism and order in the sequence of reactions a given cofactor will act by analyzing how two cofactors affect the Amax and EC50.

Deducing the Temporal Order of Cofactor Function in Ligand-Regulated Gene Transcription: Theory and Experimental Verification

Edward J. Dougherty, Chunhua Guo, S. Stoney Simons Jr, Carson C. Chow

Abstract: Cofactors are intimately involved in steroid-regulated gene expression. Two critical questions are (1) the steps at which cofactors exert their biological activities and (2) the nature of that activity. Here we show that a new mathematical theory of steroid hormone action can be used to deduce the kinetic properties and reaction sequence position for the functioning of any two cofactors relative to a concentration limiting step (CLS) and to each other. The predictions of the theory, which can be applied using graphical methods similar to those of enzyme kinetics, are validated by obtaining internally consistent data for pair-wise analyses of three cofactors (TIF2, sSMRT, and NCoR) in U2OS cells. The analysis of TIF2 and sSMRT actions on GR-induction of an endogenous gene gave results identical to those with an exogenous reporter. Thus new tools to determine previously unobtainable information about the nature and position of cofactor action in any process displaying first-order Hill plot kinetics are now available.

Are Strads overrated?

In classical music, there is a mystique surrounding Seventeenth Century violins made in Cremona, Italy and especially the Stradivarius.  These violins can cost millions of dollars and are supposed to be unmatched in sound quality by any violin made since.  People have speculated that it is the wood, the glue, the varnish or some mysterious unknown quantity that makes them so much better although nothing has ever been pinpointed.  Now, a study recently published in PNAS (see here) finds  that the superiority of the Stradivarius may be more myth than substance.  The study found that top-level violinists preferred modern violins to the classic Cremonese ones.  It was the first every study that was double blinded so that neither the violinist nor tester knew which violin was being played.  It is well-known in psychology that people’s preferences are strongly influenced by context.  An example, is that wines perform better in taste tests when they are believed to be more expensive.  The study has been criticized in that it was done in a hotel room and not on a concert stage.  I’m sure a followup is in the works.

Metaphysics as mathematics

One of the branches of western philosophy is metaphysics, which asks about the nature of being and the world.  It is the extension of what was once known as natural philosophy.  Modern science is empirical  natural philosophy.  Instead of trying to answer questions about how the world is the way it is by thinking about it, it makes hypotheses and tests them experimentally or observationally.  The late twentieth century was a time when physics, specifically string theory, drifted back towards metaphysics.  String theorists attempt to answer questions about our reality by constructing theories that are mostly grounded on mathematically aesthetic principles.   I have no real problem with string theory per se, except in its claim to be more “fundamental” than other branches of physics.  As I have argued before (e.g. here), there are fundamental concepts at all energy and length scales.

What I will argue here is that we have been misguided in trying to reunite metaphysics with science.  As I have argued before (e.g. here  and here), it is not even simple to define what is meant by “fundamental laws” or a “theory of everything”.  If our universe can be approximated arbitrarily accurately by a computable one (yes I know some of you disagree with this assertion), then what constitutes the underlying theory?  Is it the program that generates the universe?  Is it the most simple description (in which case it is not computable)?  Or is it something else?

While metaphysics as science is a dead-end for me, metaphysics as mathematics is ripe for very interesting insights. Instead of asking directly about “our” reality, we should be asking about hypothetical realities.  We should be doing philosophy of science and metaphysics on artificial worlds.  This would then be a controlled situation.  Instead of speculating about the underlying laws of our universe, we can simply specify a given set of properties in some hypothetical or simulated universe and probe the consequences.  We can do this at arbitrary levels as well –  universe,  multiverse, meta-multiverse and so forth.

I think ironically that doing such a thing would give more  insights into our universe than what we are doing now.  For example, if we started to investigate what types of simulated worlds would generate life, it may inform us more about how probable life exists in our universe ( as well as force us to come up with some quantitative definitions for life) then sending out space probes (e.g. see here).  It could also give us an idea of how variable life can be.  We seem to be stuck on looking for biochemical life.  Well maybe there are electromagnetic plasma life forms out there.  If all it took to generate complex life-like objects was a nonlinear rule that didn’t blow up, then the answer to why our universe seems so well-tuned for us would be that any old rule would have worked although it would give entirely different looking life forms.  Also, if we thought more about how we could generate or detect any type of consciousness in a simulation, that may help us better understand the consciousness we have.

A new model for publishing

Two months ago in a guest editorial for DSWeb (see here), I expressed some dismay that while we have had great inovation in many aspects of our work lives, the current (broken) publication model has remained relatively unchanged.  Now my colleagues at NIH – Dwight Kravitz and Chris Baker have published a stimulating and provocative article (see here) highlighting the many problems with the current situation, especially with the wasteful treadmill of trying to get something into a “high impact” journal, and propose a new model.  Although this will mostly have salience for people in fields that try to publish in journals like Nature and Science, I recommend that anyone who publishes should read the paper and form their own opinion.  Here is mathematician Kreso Josic’s take on the paper.

From my view as a physicist cum mathematician cum biologist, I’ve seen publishing from several perspectives.  The theoretical physics/applied math world seems to have a good system already in place where everyone posts their papers on the arXiv and then publish in an “obvious” physics or math journal like one of the Physical Review or SIAM ones.  These journals are fairly low cost for the authors, if you don’t want colour figures or physical preprints (but not cheap), and they have a nice system of transferring to sister journals if you are rejected automatically so the review process is efficient.  However, publishing in the biology world is more of a nightmare that is well documented by Dwight and Chris in their paper.  Here, getting into a high impact journal like Nature or Science can make or break your career and the chances of getting in are slim.  Authors spend a lot of their time and energy trying to get their work published and if you have little name recognition in a field it is extremely difficult just to get your paper reviewed by the more prestigious journals. Dwight and Chris have some excellent ideas of how to fix this system, which  I think have a lot of merit.  The one thing I would like to see is to make the cost for authors be as low as possible so that it doesn’t impede low funded labs.