World’s Most Wired
Biophysicist
Erez Lieberman Aiden
Erez Lieberman Aiden leads a group meeting for the Hi-C 3-D genome modeling project in Pierce Hall at Harvard University. Photo: Noah Devereaux/Wired
Erez Lieberman Aiden wants to make a point about the fractal topology of chromosomes, so he’s sitting on the floor of a CVS grocery aisle in Harvard Square, looking for the right kind of ramen noodle.
Ramen noodles, it turns out, bear a certain resemblance to the stuff of life, though not any ramen will do. None of the fancy brands, nothing pre-cooked, just undergraduate dietary staple Nissin Top Ramen (Oodles of Noodles™) in an orange plastic wrapper.
An hour later, over a simmering pot, Aiden explains that ramen comes in an unentangled state: Drop a block into hot water, let it soften, and you can fork out strands without disturbing other noodles. Bring the water to a boil, though, let the noodles churn, and you can’t take a forkful without dragging along the whole mass.
“Being unentangled is entropically unfavorable,” he says. “Equilibrium for a long chain is to be knotted.” Practical enough to complain that a packet of noodles now costs 67 cents, Aiden can’t help talking like the 32-year-old wunderkind computational biophysicist he is.
Put another way, the noodles want to be tangled. It’s their natural, default state, the path of least resistance. Given the chance, they knot up. This is what your chromosomes, those long noodly spools of DNA and protein coiled inside every cell — so often depicted in X-shaped neatness by biology textbooks and commercial iconography — should do, too. But somehow they don’t.
Instead, chromosomes take what’s called a “fractal globule” form, twisting and looping into a state that’s extraordinarily dense, yet completely unentangled. With a small enough fork, you could pull part of one chromosome out without disturbing the rest.
It’s a marvelous shape, this fractal globule, and may even prove integral to life. Yet until 2009, when Aiden and colleagues unveiled Hi-C, a new technique for reverse-engineering the structure of entire chromosomes, the form had only been glimpsed in shadowy pieces.
With support from the National Institutes of Health, Aiden is now studying chromosome structure at even higher resolution. He wants to learn how genomic form relates to function, perhaps helping to explain central questions of disease and development that remain largely unanswered.
“There’s been this humongous black box: What exactly is the role that structure plays in the genome?” Aiden said. “Now we can start looking at that.”
“There’s been this humongous black box: ‘What exactly is the role that
structure plays in the genome?’ Now we can start looking.”
— Lieberman Aiden
— Lieberman Aiden
Still just a graduate student when Hi-C and his fractal globule graced the prestigious cover of Science, Aiden is now a fellow at Harvard. In the intervening years, dozens of researchers have used Hi-C and its methodological descendants to open that genomic black box for themselves. In a way, Aiden’s legacy is as much about the method, the big-picture technique, as what he found.
That’s only appropriate. Aiden has always been a big-picture thinker; raised in the Midwood section of Brooklyn, it’s hard not to hear him without thinking of the protagonist of Pi, Darren Aronofsky’s film about a Brooklyn mathematician’s search for life’s underlying pattern, albeit without the character’s angst and pathologies.
“I enjoyed mathematics from a very young age,” he said. “At the beginning of college, I had this illusion, which was kind of silly in retrospect, that if I just understood math and physics and philosophy, I could figure out everything else from first principles.”
To be fair, neither does Aiden share the Pi protagonist’s archetypal scrawniness. A sturdy six feet tall, he has the slight paunch and strong back of a man who works long desk hours and spends a lot of time carrying infants. In June, his 2-year-old son was joined by a newborn daughter.
Aiden attended Princeton University, where his attentions ranged from the mathematics of linguistic development to the implications of German philosopher Ludwig Wittgenstein’s later statements on the nature of knowledge. “I had a wonderful time,” Aiden recalled.
Amidst these heady pleasures, however, he realized that “abstraction was not what I wanted to be doing. I wanted to do things with a tangible impact on people in my lifetime.” After graduating in 2002, Aiden spent a year at Yeshiva University — his extended family are Hasidic Jews — with the intent of describing historical change with mathematical equations, but his heart had already taken a comparatively pragmatic turn.
In fall 2003 he started a Ph.D. at Harvard and MIT, jointly supervised by evolutionary dynamicist Martin Nowak and geneticist Eric Lander. Both are famous in their fields: Nowak for research on evolution’s basic rules and the non-biological evolution of culture, and Lander as a Human Genome Project leader and interpreter of genomes. In their new student, their thinking cross-pollinated.
“It wasn’t totally unnatural to think about cultural evolution, and once I started, it became natural to say, ‘What’s going to be really transformative in this space?’” Aiden said. “It would be the ability to bring massive quantities of information to bear, based on the genomic model.”
Continue reading
Raw Materials
- Top:
- A 3-D genome map of chromosome 14. Each pixel represents the similarity between the spatial neighborhoods of a pair of megabases. (One megabase equals one million letters of DNA.) Position on the x- and y- axes represent their position on the chromosome. (Credit: Erez Lieberman Aiden)
- Bottom:
- A streamlined three-dimensional rendering of a Hilbert curve, the fractal space-filling curve into which chromosomes fold without tangling. (Credit: Miriam Huntley, Rob Scharein and Erez Lieberman Aiden.)
Erez Lieberman Aiden leads a group meeting for the Hi-C 3-D genome modeling project in Pierce Hall at Harvard University. Photo: Noah Devereaux/Wired
W
orking with fellow Harvard polymath Jean-Baptiste Michel, a systems biologist and psychologist who also studied under Nowak, Aiden scanned hundreds of library books, building up a massive, millennium-spanning database of the English language. Out of this came a high-profile Nature paper on the evolution of verb conjugation, the seeming dryness of which belies its landmark status. For decades, the idea of cultural evolution had intrigued scientists, but never had it been studied in such exhaustive, quantitative detail.
The research was also exhausting. The researchers might have stopped altogether, but for a curious observation made during their 18 long months of book-scanning: Especially for old, obscure titles, the last person to take the book out often wasn’t a person at all, but Google, which had just started its Google Books project.
Elana Stamenova and Erez Lieberman Aiden talk while walking through the Harvard campus following a weekly meeting for the Hi-C 3-D genome modeling project. Photo: Noah Devereaux/Wired
7 Favorite Paradoxes
1
Berry’s Paradox: ‘The smallest number that can’t be defined in 12 words or less’ (is a
definition in 12 words or less)
2
Braess’ Paradox: It is possible to build a new road in such a way that it takes longer
for everyone to get where they want to go.
3
Arrow’s Paradox: It is hard to design a good voting system!
4
Banach-Tarski Paradox: ‘A ball can be chopped up and reassembled into two balls.’
5
Exchange Paradox: There are two identical envelopes, one of which contains twice as much money as they other. You pick one at random, but before you open it, you are given the opportunity to switch. Should you? If so, should you switch twice?
6
Raven Paradox: ‘All ravens are black’ is logically equivalent to ‘all non-black objects are not ravens.’ So if you see a pink elephant (it’s not black and not a raven), is this evidence that ‘all ravens are black’?
7
Goodman’s Paradox: An object is called grue (resp. bleen) if it is green (resp. blue) before 2020 and blue (resp. green) afterward. [Of course, an object is called blue (resp. green) if it is bleen (resp. grue) before 2020 and grue (resp. bleen) afterward.] Is it just as logical to expect that grue objects will remain grue on Jan. 1, 2020 as to expect that green objects will remain green?
Aiden cold-emailed Peter Norvig, Google’s research director, proposing a marriage of their cultural evolution-analyzing techniques and the company’s massive new database. This union became Google Ngrams, a tool for studying cultural trends across 500 years and 500 billion words.
Unlike Hi-C, insights to emerge from Ngrams are so far more interesting than profound. (Among the observations made in the Science paper announcing Ngrams: Fame faded faster in the 20th century than the 19th, while technological innovations were adopted ever more rapidly.) But like Hi-C, Ngrams is a platform, a big-picture tool for enabling further research.
“I have a huge admiration for Erez. He’s done a wonderful job developing new tools for reading this vast database,” says Anthony Grafton, a Princeton University historian and former president of the American Historical Association.
“The interesting question is: If you have some vast amount of human culture in its written form, accessible to mining and analysis, can you make predictions from it?” Grafton continues. “It’s an open-ended question, and I’m really interested to see how it’s answered.”
Even as Aiden worked on the algorithmic engine of Ngrams, another challenge hovered in the back of his mind. Years earlier at Princeton, a professor had described the difficulty of visualizing what happens inside cells, especially at molecular levels. “It’s so hard to figure out what’s going on in biological systems. You just can’t see them,” Aiden says. In 2007, seeking a specific problem to solve, he learned of the difficulty of mapping genome structure.
Because chromosomes are so densely arranged as to be impenetrable by electron microscope, and gene sequencing destroys their physical shape, researchers relied on an ingenious but tedious workaround: chemically freezing chromosomes, then breaking them into millions of pieces from which the original three-dimensional arrangement could be inferred. For even a small portion of a single chromosome, solving this origami jigsaw puzzle gone wild required months of work.
For Aiden, the proverbial light bulb went off. Using different chemicals, he realized, would allow those millions of pieces to be analyzed far more rapidly than before. It would even be possible to map not just a chromosome segment, but the entire structure. “I thought it would be fun to scale it up,” he recalls. His optimism was not widely shared.
“People thought it couldn’t be done. It was too large a problem,” said Job Dekker, a genome structure specialist at the University of Massachusetts Medical School whom Aiden approached with the idea that became Hi-C. “With Hi-C, people realized that you could actually do this. It’s big, it’s powerful, and you start learning things about gene regulation.”
Using Hi-C, Aiden and collaborators discovered the fractal globule structure he so helpfully illustrated with ramen noodles, though the analogy soon frays. What they found wasn’t just an orderly clump. It was a structure of truly beautiful complexity, so completely filling three-dimensional space that the chromosomes inside a single sea slug neuron, which laid end to end would stretch the length of Long Island, have a volume of one cubic millimeter — and all, of course, without a single tangle.
“If you think of that genome in terms of the information content of its bases, it’s 200 petabytes of information,” Aiden said. “It’s a degree of compactness that’s completely insane, even by the standards of modern data compression. There’s nothing like it.”
What is the purpose of this extraordinary shape? Aiden doesn’t know. He suspects, as do many researchers, that it’s somehow linked to the symphony of protein production and gene modulation that occurs constantly in every living cell. With the benefit of three years’ refinement — Hi-C’s resolution is now 100 times greater, he claims — he’s now trying to link function with topography.
“This is a new kind of frontier,” he said. But for Aiden, there’s little time on the frontier for noodles. “There was a period in my life when I was eating ramen non-stop,” he says. “These days, less so. Once you have a kid, you end up eating a lot of foods with broccoli in them. You have to get your fractals other ways.”
World’s Most Wired
Wired is putting a spotlight on the brightest geniuses you’ve never heard of — the entrepreneurs, scientists, artists and designers who are quietly shaping the future behind the scenes. They’re the World’s Most Wired, and we’ll be profiling one of them bi-weekly through the end of the year. Check it out here.
DIGITAL JUICE
No comments:
Post a Comment
Thank's!