Encoding the Brain

Before he was a neuroscientist, Jason Bohland built artificial brains. While studying for his master’s degree in electrical engineering, he designed computer models that simulated the encoding and storage of memories. “I was working with simple units meant to mimic neurons,” he says, “and looking at how the connections among them can affect the capacity of memories, and how the dynamics of that artificial system unfold over time.”

Bohland (GRS’07), an assistant professor in BU Sargent College’s Department of Health Sciences, has since turned his attention to real brains and now is researching the role of genetics in certain deficits associated with developmental disorders. Why does delayed spoken language affect only half of people diagnosed with autism spectrum disorder? What are the underlying relationships among genetics, brain architecture, and behavior? Answering such questions could eventually lead to the generation of improved diagnostic and therapeutic tools.

Yet Bohland has not entirely left his early engineering work behind. He is still looking at the structure and dynamics of whole systems, using computer analysis to explore how the elements of the brain work together. He says, “My engineering experience allows me to look at the system we’re studying and ask, ‘How is it put together? If I were to build a system to do the things a brain has to do, how would I go about doing it?’”

Bohland is part of a multidisciplinary vanguard in neuroscience that combines the holistic perspective of systems biology with new computational capabilities—in this case, the analysis of huge data sets, involving terabytes of information—to revisit long-standing questions about the brain. “Historically, because of technological limitations, researchers have generally spent their careers working on one part of the brain, using a set of techniques of their choice, which may or may not be the same set of techniques another lab uses in another part of the brain,” Bohland says. “While it has led to a lot of great insights, we’re left without many big-picture ideas about neuroscience.”

“My engineering experience allows me to look at the system we’re studying and ask, ‘How is it put together? If I were to build a system to do the things a brain has to do, how would i go about doing it?’”

But technological advances in the last decade have exponentially increased the capacity to store, analyze, and share data, opening a new front in the quest to understand the workings of the brain. At Seattle’s Allen Institute for Brain Science, for example, scientists have built a standardized atlas of gene expression for the entire mouse brain and are finishing work on a similar atlas of the human brain. Both atlases are available online to any researchers interested in working with the data.

Bohland worked with the mouse data as a postdoctoral fellow and scientific informatics manager at Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, an elite institute for the study of molecular biology and genetics. Now, at Sargent College, he plans to use the forthcoming human brain data for a project that he hopes will shed light on the genetic components of certain heritable disorders affecting speech and language, and to begin to account for the wide range of behavioral variability in individuals with these disorders.

The study will engage healthy adults and children in tasks selected to highlight certain behaviors that are impaired in people with autism, specific language impairment, and other conditions that affect speech and language. With fMRI (functional Magnetic Resonance Imaging), he will localize the brain areas that are activated, and the functional interactions between these areas, as the subjects carry out their tasks. Then, using advanced computational techniques, he will compare his findings to the Allen Institute data to identify genes or sets of genes that tend to be highly expressed in those brain systems. “This provides a way to bridge the genotype-phenotype gap: We can get from the genes to the systems they’re expressed in, and we can get from behavioral outcome to the parts of the brain associated with that behavior,” Bohland says.

As with much systems-biology research, this work is data-driven rather than hypothesis-driven. But Bohland describes his work as a “hypothesis generator”: He hopes his study of healthy subjects will help him formulate theories about the mechanisms of such disorders as autism, dyslexia, and stuttering. “Jay’s research has the potential to provide new insights into the neurocomputational underpinnings of a large number of genetic disorders,” says Professor Frank Guenther of the Departments of Speech, Language & Hearing Sciences and Cognitive & Neural Systems. “This knowledge will be valuable in guiding pharmacological as well as behavioral treatments for these disorders.”

Bohland intends for the tools and methods he develops to be available for use by other researchers so that they may use them to develop and test their own hypotheses. A planned online portal will provide access to his imaging results and other resources. “More and more people are adopting the spirit of ‘We’re all in this together,’” he says. “In terms of science, that spirit of sharing data and tools is a huge component of what I believe in.”

A version of this article was published in the fall 2010 issue of Inside Sargent.