May 1, 2015 (Vol. 35, No. 9)
Alex Philippidis Senior News Editor Genetic Engineering & Biotechnology News
BigNeuron, BRAIN, and Europe’s HBP Aim to Find the Inner Workings of the Mind
The human brain consists of about 100 billion neurons, yet no one knows how many distinct types or shapes of those neurons exist. Creating a framework for answering that basic question and many other, much more complex questions is among the goals of a new multi-institutional research effort.
The project, called BigNeuron, aims to create reliable high-throughput and quantitative 3D reconstructions of the thousands of branches that make up individual neurons. It’s a crucial step, say researchers, to ultimately understanding how the brain encodes information. To create the most reliable reconstructions, BigNeuron’s dozen mostly institutional partners will develop standards for defining the best algorithms and methods, then “bench-test” those algorithms on large-scale datasets.
Hanchuan Peng, Ph.D., lead organizer of BigNeuron and associate investigator at the Allen Institute for Brain Science, which announced the project, told GEN the neuron reconstruction algorithms will be bench-tested by being ported onto a common platform so that they can be run with the same raw image input format and output reconstruction file format, in a parallel and unbiased way.
“The bench testing will be done not only for a few images, but will be done against as many as single neuron image datasets contributed worldwide from many research groups, for many different species as well as different brain regions. The analyses will be done jointly by the community,” Dr. Peng said.
BigNeuron intends to produce a large set of open-source, community-based tools for neuroscience studies, as well as standardized protocols for researchers to create their own neuron morphologies: “We’ll focus on neuron reconstruction algorithms as well as neuron comparison and search tools. In brain research, these methods may play similar roles to gene sequencing techniques and sequence comparison/search tools for genome research,” Dr. Peng added.
Also envisioned among results by BigNeuron’s partners are standardized protocols for researchers to create their own neuron morphologies, and a rich library of morphological feature definitions and algorithms to provide a foundation of quality metrics and classification organized in an openly available database of single neuron morphology data.
Lining Up Funding
BigNeuron’s launch comes as interest in researching the human brain has grown enough in recent years to spawn two large-scale, public-private research efforts.
In the U.S., the Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative would see its funding balloon in President Barack Obama’s proposed budget for the 2016 federal fiscal year—to more than $300 million, from $200 million in the current fiscal year, which ends September 30. About half of BRAIN’s FY 2016 funding would come from NIH at $150.7 million, nearly double this year’s $81.4 million in FY 2015.
How much funding BRAIN will actually see next year, and from what agency, depends on Congress. The chair of the House Labor-HHS Appropriations Subcommittee, Rep. Tom Cole (R-OK), last month warned NIH Director Francis S. Collins, M.D., Ph.D., that his agency can expect less than Obama proposed.
NIH hopes to step up activity within and across four partner federal agencies in BRAIN—FDA, NSF, the Defense Advanced Research Projects Agency (DARPA), and Intelligence Advanced Research Projects Activity (IARPA). Cross-agency efforts can be expected along the lines of the Collaborative Research in Computational Neuroscience (CRCNS) program. NIH and NSF have joined with the German and French funding agencies and the United States-Israel Binational Science Foundation to review and fund proposals intended to add to “understanding of nervous system structure and function, mechanisms underlying nervous system disorders, and computational strategies used by the nervous system.”
CRCNS awards an estimated 15 to 25 grants annually, with between $5 million and $20 million expected to be available per year depending on the quality of applications and availability of funds. This year’s proposal deadline is October 29.
“Once applications come in, they’re reviewed at NSF, but NIH participates in the review, and we pick up some of the applications and pay for them. I could foresee a similar type program going forward” as part of BRAIN, Michelle Freund, Ph.D., a project officer with NIMH, told GEN.
The number of BRAIN Initiative-related funding opportunities has grown to 13 this fiscal year from four in FY 2014. One opportunity has been approved to date for FY 2016—a grant program to support neuroscience research arising from invasive surgical procedures that provide the ability to record and stimulate neurons within precisely localized brain structures in humans.
“Applications can span the spectrum from experimental studies of mechanisms of human sensory-motor, perceptual, cognitive, mnemonic, affective, and motivational processes, to disorders of the human nervous system, to studies of mechanisms of action of device neuromodulation therapies,” according to NIH.
HBP: Rethinking the Science
So far BRAIN has avoided the sort of public discontent from researchers seen in the European Commission’s Human Brain Project (HBP). “Since their project is focused on computation approaches and ours is not, there is a natural collaboration,” Dr. Freund said, adding that leaders of both initiatives keep in regular contact. “They will need large amounts of data measured in U.S. labs (and elsewhere) for their model. We in turn should be able to use their models as they become available.”
HBP’s board on March 18 approved recommendations for reworking the project included in a 53-page Mediation Report crafted by a committee headed by Wolfgang Marquardt, Ph.D., chairman of Germany’s Forschungszentrum Jülich research center. The panel was convened after some 200 scientists threatened to boycott the HBP, contending that the 10-year, €1.2 billion ($1.3 billion) mega-initiative was not being properly managed and thus would not fulfill its goal of simulating the inner workings of the human brain.
The report urged HBP’s leaders to rethink the project’s science. While not ruling out a brain simulation, the report instead portrayed HBP as evolving long-term into Europe’s hub for simulation-based and computational neuroscience, with fewer, smaller projects that will be easier to fund and accomplish shorter-term.
“Research activities should focus on the development of a set of models that complement each other and integrate multiple scales and perspectives, together with the specification, design, implementation, and testing of IT platforms enabling and exploiting these models,” the committee concluded.
The committee recommended that HBP should re-integrate cognitive and systems neuroscience through new projects that cut across, and thereby link, existing subprojects, at an estimated cost of €45 million (about $49 million). That’s a no-brainer, since eliminating cognitive and systems neuroscience from HBP’s second funding round helped touch off the researcher revolt.
The panel also endorsed a governance change that was carried out earlier when HBP’s three-person executive committee was disbanded. The report urged a shift in oversight for HBP from the École Polytechnique Fédérale de Lausanne (EFPL), whose professor Henry Markram is HBP’s leader as coordinating scientist, to “a new legal entity jointly represented by those institutions that most strongly contribute to the project.” That entity, the report added, should be a permanent international institution, akin to the European Molecular Biology Laboratory.
Two longtime critics of HBP view the report’s findings favorably, while skeptical on how it will be followed up.
Peter Dayan, Ph.D., director of the Gatsby Computational Neuroscience Unit at University College London and a Mediation Committee member, told GEN that HBP should focus instead on developing tools that neuroscience researchers would adopt in their work.
“That should be, in the end, the mark of success for this project. That’s a huge task. At the moment, they haven’t done that yet,” Dr. Dayan said. “The idea of focusing on generating tools which could be widely used would be great. But of course, the proof of the pudding is in the eating. Are these tools usable, and are they used? If they’re not, then that’s a mark of failure.”
Alexandre Pouget, Ph.D., group leader of the Laboratory of Cognitive Computational Neuroscience at the University of Geneva, told GEN he’s waiting to see how the recommendations are translated later this year into changes to the HBP's research and governance roadmap, the Framework Partnership Agreement. He was excluded from the committee: “We’ll know when the FPA is published how much of the recommendations they’re really implementing, so we’ll have a better picture in about four or five months from now.”
The HBP would be foolish to delay serious changes to its science. One source of ideas could be the complementary brain research efforts occurring outside Europe. The report offers as one example the diffusion tensor imaging data of connectivity from the BRAIN initiative: “The emerging interaction of the HBP with these and other initiatives is encouraged to develop and exploit complementary scientific expertise, with a focus on data acquisition and knowledge generation to leverage the work of the international community.” As academia and industry have found out in recent years, partnering not only offers the promise of positive results, it’s less expensive too.