Edit: Also check out the story by the Washington Post.

Shirley is a collaborator of mine who works on using gene expression data to get a better understanding of ovarian cancer. She has a remarkable personal story that is featured in a podcast about our work together. I laughed, I cried, I can’t recommend it enough. It can be found on itunes and on soundcloud (link below).

As a physicist, I’m drawn towards simple principles that can explain phenomena that look complex. In biology, on the other hand, explanations tend to be messy and complicated. My recent work has really revolved around trying to use information theory to cut through messy data to discover the strongest signals. My work with Shirley applies this idea to gene expression data for patients with ovarian cancer. Thanks to Shirley’s amazing work, we were able to find a ton of interesting biological signals that could potentially have a real impact on treating this deadly disease. You can see a preprint of our work here.

I want to share one quick result. People often judge clusters discovered in gene expression data based on how well they recover known biological signals. The plot below shows how well our method (CorEx) does compared to a standard method (k-means) and a very popular method in the literature (hierarchical clustering). We are doing a much better job of finding biologically meaningful clusters (at least according to gene ontology databases), and this is very useful for connecting our discovery of hidden factors that affect long-term survival to new drugs that might be useful for treating ovarian cancer.

TCGA clusters

 

 


Here’s one way to solve a problem. (1) Visualize what a good solution would look like. (2) Quantify what makes that solution “good”. (3) Search over all potentials solutions for one that optimizes the goodness.

I like working on this whole pipeline, but I have come to the realization that I have been spending too much time on (3). What if there were a easy, general, powerful framework for doing (3) that would work pretty well most of the time? That’s really what tensorflow is. In most cases, I could spend some time engineering a task-specific optimizer that will be better, but this is really premature optimization of my optimization and, as Knuth famously said: “About 97% of the time, premature optimization is the root of all evil”.The docker whale



Abstract starfish

This one is just for fun. There’s no deeper meaning, just a failed experiment that resulted in some cool looking pictures.

Abstract bear


 

You have just eaten the most delicious soup of your life. You beg the cook for a recipe, but soup makers are notoriously secretive and soup recipes are traditionally only passed on to the eldest heir. Surreptitiously and with extreme caution, you pour some soup into a hidden soup compartment in your pocket.

The Information Sieve

When you get back to your mad laboratory, you begin reverse engineering the soup using an elaborate set of sieves. You pour the soup through the first sieve which has very large holes. “Eureka! The first ingredient is an entire steak.” Pleased with yourself, you continue by pouring the soup through the next sieve with slightly smaller holes. “Mushrooms, of course!” You continue to an even smaller sieve, “Peppers, I knew it!”. Since it is not a just a laboratory, but a mad laboratory, you even have a set of molecular sieves that can separate the liquid ingredients so that you are able to tell exactly how much salt and water are in the soup. You publish the soup recipe on your blog and the tight-lipped chef is ruined and his family’s legacy is destroyed. “This is for the greater good,” you say to yourself, somberly, “Information wants to be free.”

This story is the allegorical view of my latest paper, “The Information Sieve“, which I’ll present at ICML this summer (and the code is here). Like soup, most data is a mix of different things and we’d really like to identify the main ingredients. The sieve tries to pull out the main ingredient first. In this case, the main ingredient is the factor that explains most of the relationships in the data. After we’ve removed this ingredient, we run it through the sieve again, identifying successively more subtle ingredients. At the end, we’ve explained all the relationships in the data in terms of a (hopefully) small number of ingredients. The surprising things are the following:

  1. We can actually reconstruct the “most informative factor”!
  2. After we have identified it, we can say what it means to “take it out”, leaving the “remainder information” intact.
  3. The third surprise is negative: for discrete data, this process is not particularly practical (because of the difficulty of constructing remainder information). However, an exciting sequel will appear soon showing that this is actually very practical and useful for continuous data.

Update: The continuous version is finally out and is much more practical and useful. A longer post on that will follow.


The Bandwagon

Shannon’s birthday has passed, but I thought I would jump on the bandwagon late, as usual. Shannon himself recognized that information theory was so compelling that it encouraged over-use. He wrote an article saying as much way back in the 50’s.

It will be all too easy for our somewhat artificial prosperity to collapse overnight when it is realized that the use of a few exciting words like information, entropy, redundancy do not solve all of our problems”

-Shannon

As a researcher who tries to use information theory far beyond the domains for which it was intended, I take this note of caution seriously. Information as defined in Shannon’s theory has quite a narrow (but powerful!) focus on communication between two parties, A and B. When we try apply information theory to gene expression, neuroimaging, or language, we have many, many variables and there is not an obvious or unique sense of what A and B should be. We don’t really have a complete theory of information for many variable systems, but I think that is where this bandwagon is headed.


Alien?

As a child, I was visited by an alien. I remember the sensation of not being able to move or speak and seeing this other-worldly face. Some time later, when I saw a documentary about people who had been visited by aliens, I felt a chill of recognition. Their experiences matched my own.

People all over the world describe the appearance of aliens in a similar way, and this is often cited as proof that they are among us. I think that there is a different and more plausible explanation for the universality of this phenomena.

The alien image above was generated automatically from a collection of human faces. How was that done? Basically, I train some “neurons” to capture as much information about human faces as possible. These neurons try to split up the work in an optimal way, and robustly we see a neuron that represents the alien face. When you add this together with a few other informative facial features, you can flexibly recognize many types of faces. Seeing this pattern causes a strong sense of recognition because it activates core features in our facial recognition circuitry. However, having this neuron fire by itself is unnatural – no human face would cause just this one pattern to fire without any accompanying ones. Therefore it also strikes us as alien.

Dreams and drugs cause random firing patterns that sometimes activate unusual combinations of our facial recognition circuitry. The universality of people’s perceptions of alien faces only reflects the universal principles underlying the circuits in our head. Now we just have to reconstruct these principles: the truth is out there (in here?).