I has new addiction?

Eduroam is finally letting me play Phylo, a game about aligning DNA sequences. (Woo-hoo! The things that make me happy…) In the game, they are made of coloured blocks, not letters, and there’s an annoying background music you could theoretically turn off but it’s kind of part of the experience. And I can’t. Stop. Playing.

It’s quite interesting, though, in a way.

Things I’ve learnt so far:

1. I have no patience for this shit.

2. Consequently, I suck at DNA sequence alignment.

3. But I’m still better than a computer.

4. Rodents are weird.

It’s a little scary how often and how easily I can beat the computer’s score when I’m not even seriously trying. And there’s not enough excess human brainpower in the world to do all the aligning that needs doing. I suppose the problem is that there isn’t enough computer power either, which is why programmers have to settle for such shitty algorithms in the first place.

I wonder how the same game would look with protein sequences. I guess colour-coding 20 different amino acids would be a little tougher than four bases…

(Hey, idea. Someone should make a game out of Hox gene classification. Wonder if that would solve the problem I spent my undergrad project trying and failing to solve…)

Why citizen science is great

Make science into a game, and you’ll not only entertain thousands of people, but may also solve some of the toughest problems in your field.

Algorithm discovery by protein folding game players
Firas KhatibaSeth CooperbMichael D. TykaaKefan XubIlya Makedonb,Zoran PopovićbDavid Bakera,c,1, and Foldit Players
Foldit is a multiplayer online game in which players collaborate and compete to create accurate protein structure models. For specific hard problems, Foldit player solutions can in some cases outperform state-of-the-art computational methods. However, very little is known about how collaborative gameplay produces these results and whether Foldit player strategies can be formalized and structured so that they can be used by computers. To determine whether high performing player strategies could be collectively codified, we augmented the Foldit gameplay mechanics with tools for players to encode their folding strategies as “recipes” and to share their recipes with other players, who are able to further modify and redistribute them. Here we describe the rapid social evolution of player-developed folding algorithms that took place in the year following the introduction of these tools. Players developed over 5,400 different recipes, both by creating new algorithms and by modifying and recombining successful recipes developed by other players. The most successful recipes rapidly spread through the Foldit player population, and two of the recipes became particularly dominant. Examination of the algorithms encoded in these two recipes revealed a striking similarity to an unpublished algorithm developed by scientists over the same period. Benchmark calculations show that the new algorithm independently discovered by scientists and by Foldit players outperforms previously published methods. Thus, online scientific game frameworks have the potential not only to solve hard scientific problems, but also to discover and formalize effective new strategies and algorithms.

(The paper is open access in PNAS. Foldit has a website here.)

The shape of proteins is a really useful thing to know. If you are in drug research, it can help you find molecules that can manipulate a protein to kill a pathogen or repair a fault in a patient’s body. If you study evolution, it can help you find deep relationships among proteins that sequence similarities no longer preserve, and understand how their intricate workings evolved.

Traditionally, there have been two good ways of determining the structure of a protein. One is by purifying and crystallising it, and shooting X-rays at the crystals. Obviously, that only works with proteins that can be purified and crystallised, which is not all of them. The other is homology-based prediction – basically comparison with a related protein of known structure, which obviously requires a similar enough protein with a known structure.

Predicting the structure of a protein from its amino acid sequence is fiendishly difficult, but that’s what the folks behind Rosetta and Foldit are trying to do. Nowadays, isolating and sequencing the gene that codes for your protein of interest is relatively easy (sequencing the protein itself is tougher). If all you had to do to accurately know its structure was to plug the sequence into a program and wait a few minutes, that would make the life of many people much easier.

(I briefly played Foldit, but I admit I jut got lost and frustrated and gave up. However, I encourage everyone with an actual attention span to give it a go and see if it’s for you.)

Goin’ a-hunting

It’s a little-known fact that before/in between wanting to be a biologist, I almost got sucked into astronomy. The cosmos still fascinates me, from the menagerie of space rocks and gas balls that fill our own solar system to the mysteries at the edge of the known universe. To the evolutionist in me, the possibility of life on other worlds is an especially tantalising idea. And now we are finding other worlds at a breakneck pace. I don’t think we will ever know what life is like on any of them, though detecting its existence may once become possible.

Did I mention planet hunting is awesome?

I am talking about the citizen science project Planet Hunters, of course. This is only one of the amazing projects you can participate in at the Zooniverse (which gets its name from Galaxy Zoo, the project that started it all). The main mission of Planet Hunters is, of course, to find planets orbiting other stars. You, the user have to look at a month’s worth of brightness measurements from a star, and search for the tell-tale dips that betray an extra-solar eclipse. Like this:

Most of the more spectacular ones have already been found by this point – either by your fellow hunters, or by the team operating the Kepler space telescope, which provides all the data. However, there are so many other gems to discover among those messy light curves that it almost doesn’t matter if your planet-hunting thunder is perpetually stolen.

Sometimes, you find pure beauty. One of the most common types of Interesting Stuff that the Kepler data offer is eclipsing binaries. These are pairs of stars orbiting each other in a way that we see their orbits edge on. Like the planets, these binaries eclipse their companion stars. Since stars are bigger and brighter than planets, the eclipses are much bigger compared to the noise in the data, so an EB has neat, clean dips in its light curve, occurring with clockwork regularity.

Some of them are so close together and orbit so fast that at Kepler’s resolution, a month of their light looks more like lace than a pattern of ups and downs.

And then there are all the others; dwarfs and giants, variable stars regular and haphazard, huge flares, weird things like cataclysmic variables. Even if you are in it for the planets, you can’t help but learn a lot about the stars. After a while, they become like family. You look at a light curve and you can immediately guess whether it’s a dwarf or a giant, whether it’s cool or hot, whether it’s a binary or a loner, or even if its’s one of the rarer breeds of stars you might come across. It’s a bit like birdwatching. If you’ve ever got disproportionately excited from recognising a rare bird (or flower, or insect, or sports car), you know what I mean. (If you haven’t, what are you waiting for? ;))

I’m grateful to the people who make these adventures possible. It’s great that I can play at astronomy, see all that neat stuff, contribute to a field I have absolutely no expertise in, and learn from the knowledgeable folks that hang around the forums. The Zooniverse deserves every one of its hundreds of thousands of users and millions of clicks, is all I’m saying 🙂