Archive for November, 2012

Depending on your perspective, Twitter is either a colossal waste of time or an addictive tool that has changed the way people interact online. But there’s no arguing that the service generates a ton of data, with roughly half a billion tweets posted around the world each day. As with any enormous data set, there are likely valuable signals to be captured from within that massive noise…with the right tools. In an award-winning master’s thesis, Computation Institute scientist Mattias Lidman set out to see if he could use the avalanche of daily tweets to predict the future winners and losers of reality shows and stock markets.

Financial analysts have long looked for an information advantage in playing the market, a way of knowing and acting upon important news before other investors. Logically, people have tried to use internet activity to make predictions about future market trends, including a paper last year that looked at web search terms and trading volumes. But Twitter offers a new promising data mine from which to extract predictions, due both to its global popularity and its ephemeral nature. While writing a news article or a blog post (usually) requires both time and a computer, Twitter’s portability and brevity enables more frequent, train-of-thought use.


Read Full Post »

How HIV Builds its Suit of Armor

(L to R) The coarse-grained model of the viral capsid; Example results of lattice assembly simulations; Simulation detail showing the characteristic hexagonal lattice structure. (Image by John Grime)

At the molecular level, science is often a series of snapshots. With the most advanced imaging techniques, researchers can magnify targets over a million times, allowing them to examine structures as small as an Ångstrom. But in order to achieve this incredible resolution, most techniques require their targets to be fixed in place, reducing the dramatic, flowing motions of molecules to a series of before and after pictures.

To fill in the gaps, computational scientists such as those at the Center for Multiscale Theory and Simulation (CMTS), develop models that use these still images and the laws of physics to predict the movement of a molecule from point A to point B. In the case of a virus such as HIV, filling in those blanks could reveal potential weaknesses to exploit as new drug targets. In a new paper for Biophysical Journal, CMTS researchers John Grime and Gregory Voth simulated the intermediate steps of a critical moment for HIV: when it assembles a “suit of armor” for its genes.


Read Full Post »

Photo by Jason Smith

Computation is now an essential tool for researchers, as data analytics and complex simulations fuel ambitious new studies in the sciences and humanities. But the path from a spreadsheet on a laptop to using the world’s most powerful supercomputers can be intimidating for researchers unfamiliar with computational methods.

To help researchers along this journey, the University created the Research Computing Center (RCC), providing access to hardware and expertise to faculty and students. At an opening reception on November 8th at the Crerar Library, scientists from Argonne National Laboratory and IBM joined RCC director H. Birali Runesha in welcoming UChicago researchers to this valuable new resource.

“The mission of the Research Computing Center is to advance research and scholarship at the University,” Runesha said. “What we are trying to do here is not just to provide access to hardware, but to work with you to understand your research and integrate high-performance computing into it to achieve our major goal, which is to help you literally transform your research by performing computational analysis that would otherwise not be possible.”


Read Full Post »

Sometimes making sense of big data starts with finding the right metaphor. As cheaper genetic sequencing produces a growing waterfall of data, scientists need to organize all that new information into a form that makes the intricate world of genes and biology easier to understand, explain and investigate further. But unlike other big data fields, genetics does not have a visual “playing field” that is easy to visualize, like the night sky for astronomy or the globe for climate. Many attempts to visualize how genes interact with each other end up taking the form of the “hairball,” a dense web of points and lines that does little more than emphasize the intimidating complexity of genetic networks.

In his talk at the Computation Institute on November 9th, Mark Gerstein of Yale presented several metaphorical options for understanding the cell’s genetic network. At various points, he offered comparisons between gene interactions and a road network, a corporate organization chart or a computer operating system, each with their own strengths and weaknesses for making sense of the data. Gerstein grappled with these options as part of the ENCODE project, a large consortium formed to understand the functional elements of the human genome that recently published their first wave of articles. In order to transform understanding of the genome from a string of DNA bases to a living, interacting network of genes, regulatory elements and enzymes, ENCODE needed a new metaphor for depicting the system.

Two options appealed to the researchers. The highway system, marked as it is by well-connected hubs and bottlenecks of restricted flow, can represent the unequal distribution of connections between genes. Alternatively, a company org chart can depict the hierarchy of genetic interactions, with some regulatory elements more likely to control other factors, while some elements are more likely to be controlled by factors higher up the chain. Gerstein and ENCODE researchers took over 100 transcription factors and organized them according to the latter metaphor, finding that they fit into a three-tiered structure of “master regulators,” “middle managers,” and “workhorses.” They could then look more closely at the specific factors in each tier, finding that those elements at the top of their genetic org chart were more influential upon biological processes and more evolutionarily “conservative,” similar across different species because of their critical role for life.

A third metaphor explained by Gerstein offered insight even through how it didn’t perfectly represent the function of the genome. One way to think of a genome, he said, is as “the fundamental operating system of life,” the core instructions that run the programs of a living cell. So Gerstein compared the hierarchical network of the genome to the open-source operating system Linux, which “evolves” over time as users modify and improve its code. However, the comparison revealed that the patterns of change are very different within the two systems, both in structure (as pictured above) and the dynamics of change. While genes with large influence and many connections are resistant to evolutionary change over time, the pieces of code that are most influential in Linux are more likely to change as developers look to make improvements with the largest impact.

“In the biological situation you have random change, and you don’t want to have random changes where you have a lot of connections,” Gerstein said. “In Linux or a computing operating system, you have intelligent designers who are making changes that they believe are not going to be disruptive. Consequently, they’re going to make those changes in a very different pattern.”

Read Full Post »

Photo by Robert Kozloff.

Story by Greg Borzo for UChicagoTech.

Thirteen-year-old Gilberto is suffering from refractory neuroblastoma, a tumor of the sympathetic nervous system that has dissipated throughout his body. Recently, he hit a roadblock in his treatment that is all too common at hospitals and clinics around the country.

His doctor wanted to enroll Gilberto in a clinical trial. Alas, the trial was not yet open at the hospital where Gilberto was being treated because all of the necessary paperwork had not yet been processed.

“This kind of thing happens everywhere because of the way clinical protocols are generated and implemented,” says Samuel Volchenboum, assistant professor of pediatrics at the University of Chicago and Computation Institute Faculty.

An informatics expert as well as an oncologist, Volchenboum is out to streamline the entire way clinical trials are organized, managed, tracked and reported. To help make that possible, the Innovation Fund, the University of Chicago’s venture philanthropic proof-of-concept fund, awarded Volchenboum $40,000 in its fourth round of funding this June.


Read Full Post »