The exciting work in computational neuroscience described in this week's Science Careers and the 6 October Science magazine special issue required financial backing before researchers could get much done. So where did the money come from? And where should other computational neuroscientists turn to find support for their research? In this article, we survey several American sources of computational neuroscience research funding in the public and private-nonprofit sectors.

U.S. federal agencies have taken the lead in funding research in computational neuroscience, and the majority of government funding comes from the National Institutes of Health (NIH) and the National Science Foundation (NSF). Acknowledging the interdisciplinary nature of the field, in the past few years, these agencies have established special programs to focus their resources and coordinate their efforts. But the recent trend with computational neuroscience, at NIH and NSF, is away from high-profile special projects and toward the integration of the field into the agencies' mainstream processes.

Federal agencies

National Institutes of Health

At NIH, interest in neuroscience extends across organizational boundaries. Michael Huerta, associate director of the National Institute of Mental Health, was instrumental in starting, in 1993, the Human Brain Project (HBP), a large-scale, multi-institute initiative that combined informatics with basic research and clinical studies in neuroscience. HBP fulfilled its mission, says Huerta, and ended in 2005. "The Human Brain Project was a pioneering effort when it started 13 years ago," he says, "and since that time, the science has matured to the point where it does not need this incubator." Today, funding priorities are set by the organization's Blueprint for Neuroscience Research, which was written in 2004 to set priorities, coordinate planning, and encourage collaboration across 15 NIH institutes and centers. That Blueprint includes an emphasis on computational neuroscience and neuroinformatics.

Several current (as of October 2006) NIH program announcements provide funding sources for research in computational neuroscience; see the list below. Huerta notes, however, that unsolicited computational neuroscience proposals are becoming routine. This fact, Huerta says, is another indicator of "computational biology and biomedical informatics having become 'mainstream' research at NIH."

In the following announcements, NIH breaks out special grants for high-risk/high-impact inquiries (R21) from its regular research project grants (R01), unless otherwise noted:

Neurotechnology Research, Development, and Enhancement (R01)
Neurotechnology Research, Development, and Enhancement (R21)
The announcement for these programs says that their purpose is "to advance understanding of the nervous system and its product, behavior, through support of research, development, and enhancement of a wide range of technologies. Technologies and approaches appropriate for study, development and enhancement under this [announcement] include hardware (e.g., instruments, devices, etc.), software (e.g., computational algorithms, informatics tools, etc.), and wetware (e.g., imaging probes, genetic tools, etc.) that would be used to study the brain or behavior in basic or clinical research (or for clinical use)."

Innovations in Biomedical Computational Science and Technology (R01)
Innovations in Biomedical Computational Science and Technology (R21)
These funding programs emerged in 1999 from an NIH working group that recommended a special initiative in biomedical information science and technology. These programs target the informatics side of the field and aim to develop new information-technology tools for biomedical research. The tools being sought, according to the announcements, address issues of "data acquisition, archiving, querying, retrieval, visualization, integration and management." The programs also fund inquiries into data exchange and interoperability, analytical tools for large data sets, and new models or simulations of complex biological processes.

Collaborations with National Centers for Biomedical Computing (R01)
Exploratory Collaborations with National Centers for Biomedical Computing (R21)
A related NIH initiative is the National Centers for Biomedical Computing, a collection of interdisciplinary project teams focusing on informatics applications in the health sciences. The announcements call for proposals from individuals or small groups to work with these centers. The current group of centers, residing at universities and medical research facilities across the country, addresses fields such as systems biology, image processing, and biophysical modeling, all of which are related to computational neuroscience. They bring together computational scientists with biomedical researchers and experimental researchers. It is important to note that the announcements encourage new collaborations, which are given higher priority than funding for existing partnerships.

Bioengineering Research Grants (R01)
Exploratory/Developmental Bioengineering Research Grants (R21)
As defined by NIH's Bioengineering Consortium, bioengineering "integrates physical, chemical, mathematical, and computational sciences and engineering principles to study biology, medicine, behavior, and health. It advances fundamental concepts; creates knowledge from the molecular to the organ systems levels; and develops innovative biologics, materials, processes, implants, devices, and informatics approaches for the prevention, diagnosis, and treatment of disease, for patient rehabilitation, and for improving health."

Bioengineering Research Partnerships (R01)
These grants likewise focus on bioengineering but add building collaborations between biomedical, computational, and engineering researchers as an objective. NIH's Bioengineering Consortium has stipulated separate guidelines for these grants, including a list of priority research topics.

Continued Development and Maintenance of Software (R01)
One productive approach to software development is to enhance an existing program instead of starting from scratch. This funding program, according to its announcement, calls for work to "apply best practices and proven methods for software design, construction, and implementation to extend the applicability of existing biomedical informatics/computational biology software to a broader biomedical research community."

Mentored Quantitative Research Development Award (K25)
This program is part of NIH's career development award series and is aimed at quantitative, informatics, or engineering experts. It funds a combination of training in biomedical disciplines and supervised research experiences to build skills in the health sciences.

NIH allocates some of its R&D funding for small business (SBIR) and technology-transfer (STTR) grants. SBIR awards are made to small businesses, often in collaboration with academic labs. University PIs can apply for STTR grants in collaboration with companies. The awards applicable to computational neuroscience include:

Innovations in Biomedical Computational Science and Technology Initiative (SBIR)
Innovations in Biomedical Computational Science and Technology Initiative (STTR)
These programs apply to all NIH institutes and fund basic research in biomedical computing, as well as development of informatics tools for the health sciences. Projects must span biomedical research and computational applications. Among the biomedical research areas identified in the announcement are neurobiology and cognitive science.

Integration of Heterogeneous Data Sources (SBIR)
Integration of Heterogeneous Data Sources (STTR)
These programs address a common problem in biomedical research: the need to combine data from different sources. The problem of integrating data from various sources is not unique to the health sciences, but in an interdisciplinary field such as computational neuroscience, the problem becomes even more acute. The problems of data integration extend beyond technical factors (different equipment and software platforms) to cover semantics, legal restrictions, and social norms.

See the Science Careers article from 2004 for more information about SBIR and STTR funding.

NIH's Web site has a section on grant applications and management, which applicants should consult carefully before submission. Applicants should also get to know the review process administered by NIH's Center for Scientific Review. In computational neuroscience, many proposals may be assigned to specialized study sections, outside of the regular standing committees. Small business and technology-transfer proposals are reviewed by other specialized study sections.

National Science Foundation

In January 2006, Collaborative Research in Computational Neuroscience (CRCNS): Innovative Approaches to Science and Engineering Research on Brain Function--an $80 million collaboration among nine NIH institutes and five NSF directorates--ended its latest round of funding, and is being renewed. Kenneth Whang, program director in NSF's Division of Information and Intelligent Systems, says that NSF is considering issuing a new program announcement related to computational neuroscience, but, he says, researchers need not wait for a CRCNS renewal to find funding for their work. With computational neuroscience extending over so many disciplines, investigators can find funding from a range of NSF sources. "We have programs within computer science and biology," Whang says, "and to some extent in mathematical sciences and engineering, that supports work of a similar flavor."

Cognitive Neuroscience Program
This program seeks "highly innovative and interdisciplinary proposals aimed at advancing a rigorous understanding of how the human brain supports thought, perception, affect, action, social processes, and other aspects of cognition and behavior, including how such processes develop and change in the brain and through time." "The emphasis at NSF," the announcement says, "will be placed on integration of the cognitive sciences, social and economic sciences, and engineering in service of insights into healthy functions of brain, cognition, and behavior."

Information and Intelligent Systems: Advancing Human-Centered Computing, Information Integration and Informatics, and Robust Intelligence
The concept of robust intelligence "encompasses computational understanding and modeling of the many human and animal capabilities that demonstrate intelligence and adaptability in unstructured and uncertain environments." Research funded in this program should address issues such as "intelligent perception, communication, and reasoning capabilities that are not constrained to address only a single problem in isolation or in one particular context." Robust intelligence systems, according to the announcement, can use a variety of modeling and analytical methods, such as "analogical, statistical, and logical inference, to deal with open-ended and changing concepts and environments and to integrate possibly heterogeneous knowledge and reasoning methodologies in complementary and supplementary ways."

Biomedical Engineering (BME), Research to Aid Persons with Disabilities (RAPD), and Biophotonics Programs
This program supports research that applies engineering principles to problems in biomedical science while contributing to the body of knowledge in engineering. One part of the program, RAPD, funds technologies for development of new medical devices or software for persons with disabilities, where proposals in the field of computational neuroscience could be considered.

Joint DMS/NIGMS Initiative to Support Research in the Area of Mathematical Biology
NSF's Division of Mathematical Sciences (DMS)--part of the Directorate for Mathematical and Physical Sciences--and the National Institute of General Medical Sciences (NIGMS)--an NIH institute--seek proposals that "identify innovative mathematics or statistics needed to solve an important biological problem." Proposals should cite a direct relation between biological applications and mathematics, and research teams should include representatives from both life sciences and mathematics. Individual investigators will need to show expertise in both mathematical and biological disciplines.

Environmental and Structural Systems
This thematic area in NSF's Directorate of Biological Sciences studies "the functional evolution of organisms in their physiochemical and biotic environments. Included are studies of physiological ecology, evolutionary physiology, stress physiology, functional morphology and movement, animal sensation, computational and systems neuroscience, and environmental genomics." The announcement encourages proposals with computational or engineering approaches to biology.

If a research idea does not fit cleanly into one of these NSF solicitations or announcements, applicants can still submit a proposal. However, proposals should cite at least one NSF program area in the proposal. Applicants should review the NSF Grant Proposal Guide for details on responding to solicitations and program announcements or submitting unsolicited proposals.

Defense Advanced Research Projects Agency (DARPA)

DARPA is a Department of Defense scientific agency, which conducts its own research and funds external projects to assist DOD's mission. The agency's work covers a wide array of disciplines covering health, life, and physical sciences, as well as engineering. Its current programs include work with potential connections to computational neuroscience, such as biologically inspired robotics and smart prosthetics devices. Researchers who want to compete for these grants should keep an eye on DARPA's solicitations page for funding announcements.

Private-sector organizations

Burroughs Wellcome Fund

The Burroughs Wellcome Fund (BWF, a Science Careers sponsor) supports research that cuts across the traditional biomedical and physical-sciences boundaries. Via a program called Career Awards at the Scientific Interface (CASI), BWF encourages early-career physical scientists to apply principles from physics, chemistry, mathematics, computer science, and engineering to the biological sciences. CASI awards provide $500,000 over 5 years, supporting up to 2 years of postdoctoral training and up to 3 years of a faculty appointment. A 2004 Science Careers article explains the background and workings of CASI.

Nancy Sung, BWF's senior program officer for CASI, says that computational neuroscience fits squarely into CASI's mission, and probably five of the 36 grants awarded so far would fit under that label. "The whole point is to find people early in their career and help them establish a body of work and transition into a faculty position," she says. "We want to use our funds to do some creative work."

BWF expects to award 12 more CASI grants this year. The program is very competitive; the award rate, Sung says, runs about 10%. BWF's review process favors researchers with a record of accomplishment. "We're looking for someone who has hit two home runs, ... [who has done] very high quality work as a graduate student and postdoc. We want to see that they have been productive all along, not just in the numbers of papers they produce, but that the papers are high quality." BWF expects to announce the 2007 application cycle in November. Applicants must be nominated by their institutions, so they need to start the application process well before BWF's due date.

Swartz Foundation

Jerome (Jerry) Swartz, a bar-code-systems entrepreneur, founded the Swartz Foundation in 1994. The foundation's goal, its Web site says, is to apply "problem-solving approaches from physics, mathematics, electrical engineering and computer science into neuroscience research, with the goal of better understanding the relationship between the human brain and mind."

Hirsh Cohen, a scientific adviser to Swartz, says the foundation supports 10 research centers at universities and scientific laboratories. Cohen says that in 1999-2000, the Swartz Foundation joined with the Alfred P. Sloan Foundation to fund five research centers that Sloan had established a few years earlier--at the California Institute of Technology, Salk Institute, University of California, San Francisco, New York University, and Brandeis University. The Swartz Foundation is now the main funder of these five facilities.

Since then, the Swartz Foundation has added funding for computational neuroscience research centers at five more facilities: Cold Spring Harbor Lab, Columbia University, Princeton University, Yale University, and University of California, San Diego.

Cohen notes that most of the funding at these 10 centers goes for postdoctoral positions, where they now support 25 to 30 postdocs in computational neuroscience. "On rare occasions, we will give an individual postdoc grant for a particularly innovative unsolicited proposal," Cohen notes; still, he urges postdoc prospects to approach Swartz-funded institutions, not his foundation, for career opportunities.

McKnight Endowment Fund for Neuroscience

The McKnight Endowment Fund for Neuroscience supports research on treatment of brain injury, impairment, and disease. This fund operates separately from its parent organization, the McKnight Foundation, and has one program designed to encourage advances in technology as applied to neuroscience. This program, McKnight Technological Innovations in Neuroscience Awards, focuses on “how technology may be used or adapted to monitor, manipulate, analyze, or model brain function at any level, from the molecular to the entire organism.” The program emphasizes innovation and does not fund research on current technologies.

Awards from this fund provide up to $100,000 per year, with up to four grants per year awarded. Candidates must hold tenured or tenure-track positions at not-for-profit institutions in the United States and must be U.S. citizens or permanent residents. McKnight plans to announce the call for 2007 awards in October 2006.

The fund also has a program for early-career researchers, the McKnight Scholar Awards. This program, for candidates who have completed their postdoctoral training but with less than 4 years of subsequent experience, covers topics in general neuroscience and “seeks applicants working on problems that, if solved at the basic level, would have immediate and significant impact on clinically relevant issues.” Awards from the fund provide $75,000 for each grantee. The next deadline for applications is 2 January 2007.

Alan Kotok is managing editor of Science Careers.

Comments, suggestions? Please send your feedback to our editor.

Alan Kotok is managing editor of Careers.