INTERDISCIPLINARY/MULTIDISCIPLINARY RESEARCH

Enzymology began as a field primarily of interest to biologists who wanted information about metabolic cycles and the proteins catalyzing physiological reactions. However, as soon as relatively pure enzymes were available, organic chemists joined the fun, trying to delineate mechanisms through organic synthesis and physical chemistry principles. They were quickly joined by physical chemists and physicists, bringing in the strength of kinetics, spectroscopy, and structural biology. The advent of cloning and site-specific mutagenesis has sparked further advances in the field. Drug development based on detailed knowledge of enzymology began very early on and continues to be an important area. Enzymology was once considered a demonstration of the strength of interdisciplinary research. However, today single individuals are simultaneously carrying out research in all of the fields mentioned above, and any modern biochemistry/ molecular biology department should have all of these skills represented on its faculty.

Who would have guessed that the discovery in 1946 that nuclei can be oriented in a magnetic field would lead to modern nuclear magnetic resonance and magnetic resonance imaging? Yet this fundamental observation by physicists was rapidly developed by physical chemists and biologists into instrumentation that is indispensable for modern research in chemistry, biology, neuroscience, and psychology and that is likewise indispensable for diagnostic work in clinics. Several Nobel prizes have been awarded in this area of research.

Physicians recognized very early that deficiencies in certain substances could lead to severe health problems. Early examples included vitamins and hormones. Physiologists, biochemists, and cell biologists soon found that specific proteins mediated the mechanism of action of these substances, and this led to the field of receptor biology. Physicians, biochemists, cell biologists, and physical scientists have all contributed to the elucidation of receptor biology for such diverse substances as insulin, cholesterol, and adrenaline. Moreover, the medical implications have been very significant.

A more recent example is the sequencing of the human genome. The techniques required for this impressive accomplishment involved the collective efforts of many traditional fields, including physics, chemistry, biology, and computer science. Understanding the health implications of the sequences that have been obtained will be even more difficult and surely will involve areas such as mathematics, computer science, and bioinformatics. The identification of specific genes associated with a specific disease is an obvious health implication of this work.

Cognitive science began as an attempt to broaden the traditional accounts of behavior to account for high-level cognition such as language by reaching out to concepts and approaches from fields such as linguistics and anthropology. Soon cognitive science moved in another direction and attempted to explain behavior with neural networks, reaching out to another emerging field, neuroscience. Among other outcomes, this juxtaposition led to language models couched in neural terms. Neural net models in turn led to general algorithms for machine learning and machine classification, combining the emerging field with newly emerging trends in statistics, computer science, information science, and informatics. Applications of this synergy are now found everywhere in society.

Another good example, overlapping with the previous one, concerns the emerging fields of cognitive neuroscience and behavioral neuroscience, which represent a blend of neuroscience, functional anatomy, psychology, and physiology. These developments occurred hand in hand with the emergence of new brain imaging techniques such as positron emission spectroscopy and functional magnetic resonance imaging. The result has been an enormous growth in the understanding of mental illness and cognitive abnormalities.

When new theory building is pertinent at the intersections among disciplines, coupled with long-term programmatic changes such as those that characterize the linkages between the health and social sciences, some observers have argued for the recognition of a notion called transdisciplinary research,² contrasting it with interdisciplinary and multidisciplinary research. The notion is to provide a systematic, comprehensive framework for the definition and analysis of social, economic, political, environmental, and institutional factors that influence human health and well-being. The implications are challenging for training a new generation of individuals with a broad, integrative view of the health and social sciences.

Many major health problems faced by society are extremely complex and inherently require research from many areas of science. Examples include obesity, drug abuse, smoking, alcohol abuse, and even violent behavior. In these and other cases, inter-, multi-, and transdisciplinary research is a necessity rather than an option. It is to be noted in this regard that training in the behavioral and social sciences is only one component, but nonetheless an essential component, of the training required to deal with these health issues. At present such training largely resides in the National Institute of Mental Health. This policy places many obstacles in the way of the interdisciplinary research training that the committee regards as an essential part of the research package required for such health problems. Therefore a specific recommendation that much larger efforts be made to integrate research training in the behavioral and social sciences with research training in other fields in all the relevant institutes of the National Institutes of Health (NIH) has been included.

EMERGING FIELDS

As a result of the advancing edges of research, new fields are constantly emerging. Some come and go, whereas others develop into new, well-recognized entities. Some recent examples, in addition to those previously mentioned, are cited below.

Probably the best-recognized examples are the fields of genomics and proteomics. Both are an outgrowth of the vast number of genome sequences becoming available. Genomics is usually considered the study of DNA itself, whereas proteomics is broadly construed to represent the study of proteins expressed by genes. In both cases, regulation of genetic processes is an important factor. Work in these fields, as noted previously, requires quantitative skills in mathematics and computer sciences, along with a thorough knowledge of the associated biology.

Nanotechnology, including nanomedicine, is a closely coupled wave of the future. Already nanodevices such as DNA/RNA chips have been used to make important advances in basic research and diagnostics.

The study of biological molecules has reached the stage where single molecules can be visualized. The new techniques include fluorescence correlation spectroscopy, single-molecule fluorescence microscopy, and cryo-electron microscopy. Recent advances in nuclear magnetic resonance and X-ray crystallography have bolstered the level of information that can be obtained. This area of research will permit a new level of understanding to be reached with regard to the function and mechanism of action of biological molecules.

Impressive progress has been made in understanding the physiology of organisms by isolating the major components, such as enzymes and nucleic acids. The time is now ripe for the development of systems biology. This requires integration of the entire biological framework of an organism, starting with bacteria and ending with humans. This effort will require broad training in the basic sciences and medicine, from biology to mathematics.

Modern biological research has become nearly impossible without the use of computers. Recognition of this new reality has led in part to the emergence of a burgeoning discipline at the intersection of biology and computer science: bioinformatics. Closely related to computational biology, bioinformatics takes a computer science perspective in developing new methods and techniques pertinent to analysis of the vast amounts of data being produced by biology researchers and especially in the fields of genomics and proteomics. Bioinformatics draws heavily on methods developed by the slightly older field of clinical (medical) informatics; however, a new breed of scientist is being produced by academic programs in the evolving field of biomedical informatics (which subsumes both clinical and biological applications of informatics principles and methods).

An interdisciplinary field concerned with decision making has emerged in recent years. It is represented, for example, by the Society for Judgment and Decision Making and a related organization with a health care focus, the Society for Medical Decision Making. Scientists in this developing field come from economics, social psychology, business, law, artificial intelligence, statistics, epidemiology, anthropology, and cognitive psychology and are concerned with all aspects of decision making in the real world, both normative (what decisions are optimal) and actual (what people do). There is often a large gap between the two, and resolving the difference is of vital importance for society and health care.

Another relatively new interdisciplinary field in the social sciences combines political science, sociology, public and environmental affairs, economics, international business investment, and psychology and is concerned with rational decision making and resource management at the level of societies. This field has been supported in part by a variety of recent National Science Foundation initiatives in global change. It includes subfields such as ecological economics, environmental science, urban and rural affairs, international resource management, and much more of a similar nature.

Sometimes an interdisciplinary field is in such early stages that it has no name and there remains uncertainty about whether it will in fact emerge. Yet one cannot withhold support from such fields at early stages because support would arrive too late to do any good during the critical formative stages. One current example is in the intersection of robotics, computational models, machine learning, and developmental psychology and is concerned with behaving organisms and devices learning to operate in a real-world environment.

Currently, U.S. population demography clearly indicates a shift to older people. Hence, the study of aging will be of increasing importance. This includes not only the biology but also the psychology and sociology of life stresses, chronic medical problems, and mental health. This has already been recognized, for example, by the award in 2002 of a $26 million grant to the University of Wisconsin Institute on Aging from the National Institute of Aging. The purpose of the grant is to carry forward a project initiated in 1995 by a multidisciplinary team interested in the behavioral, psychological, and social factors of how people age. The Mind-Body Center, created at the University of Wisconsin in 1999 and funded by the National Institute of Mental Health, investigates the age profiles of physical and mental health in humans and animal models. Integrated study of aging phenomena is an important target for the future.

TRAINING IMPLICATIONS

Clearly, research in emerging areas and interdisciplinary/ multidisciplinary research are important for making major breakthroughs in health-related research. Therefore, it is important that research training be broadly based. This has been well recognized at many universities and funding agencies with the creation of programs bridging multiple departments and institutions. The Burroughs Wellcome Fund, for example, provides specific grants for the purpose of bringing students with backgrounds in the physical, computational or mathematical sciences into research in the biological sciences. Ten such programs have been funded since 1996. The National Research Service Awards (NRSAs) have long promoted broad training. The National Library of Medicine has supported training at the intersection of biomedicine and computer science for over 20 years. Training grants typically span multiple departments. In fact, this is often a requirement for a training grant.

The term “broadly based training” is not to be taken too literally. The fundamental problem with much of present-day training is caused by the increasing depth of understanding of increasingly narrow fields. However, these factors make it difficult to train any scientist truly broadly, other than in a superficial fashion. Thus, a balance should be struck in which sufficient training is provided in a discipline to allow deeper scientific progress in an established field with sufficient breadth in relevant alternative fields to allow new progress outside established boundaries. Given the impossibility of training in every field, means must be found to identify other relevant fields through which a field can be transmuted and enriched and then to encourage sufficient numbers of scientists to achieve sufficient mastery of those fields so identified. The difficulty of this task is easy to underestimate. One solution is to create “bridge” people who have sufficient breadth and understanding of two disciplines that they can help meld and catalyze communication among members of larger teams drawn from the discipline themselves. It should also be recognized that interdisciplinary training is time consuming since there is more to learn on the part of the trainees and a need for greater coordination of the trainees' research projects on the part of the mentors.

A further distinction helps delineate the concept of “broad training.” There are some skills that are so fundamental to all scientific fields and to scientific progress that all trainees must learn them; these include mathematics, quantitative approaches, statistics, computation, writing, speaking, and communication. This may seem a tautology, but all of us have seen examples of scientists who have not received reasonable training in one or more of these essential skills, and in fact there are some fields where one or another of these skills is routinely overlooked. The first and foremost recommendation is that policies be implemented that ensure all trainees receive intensive training in such areas, noting the special need for quantitative and computational training, while recognizing that potentially good science too often goes unnoticed if the researchers are unable to describe their work effectively in talks and journal articles.

The second concept of “broad training” is field-specific training that allows new concepts and approaches to develop through synergy with new areas. Examples of such synergy were provided in the introduction to this section. Identifying such areas is an ongoing enterprise that requires creativity and insight, but some general procedures and approaches can help ease the difficulties. Research supervision by mentors from more than one department should be encouraged. Emerging areas need to be quickly recognized and supported, a process that can and should occur within individual funding agencies but also can be furthered by outside groups (such as this committee) instructed to help in this enterprise. Establishing training grants in emerging areas is important, but it is a lengthy process and may be too slow to encourage the appropriate training. Individual awards, however, can respond rapidly to immediate needs, so the committees making such awards should be especially sensitive to the need to provide awards for research training proposals that move beyond traditional field boundaries. Further, the instructions to those applying for such awards should emphasize the importance of this criterion.

RECOMMENDATIONS

Recommendation 8-1: The standing committee created to monitor the continuing needs of the biomedical research community should also be charged to provide recommendations to NIH as to the identity of emerging research fields.

The need to react quickly to recognize important new research developments and to support the training of appropriate personnel is of obvious importance to the health sciences. To track the evolution of existing fields, the changes in relations among existing fields, and the emergence of new fields, both NIH and the standing committee should make use of techniques that analyze electronic databases to map existing scientific structures and their changes over time.

It is extremely difficult for individuals, no matter how knowledgeable, to grasp the structure of science and the way this structure evolves. Fields overlap in confusing ways, and existing mechanisms (funding and otherwise) often are rooted in old scientific divisions and classifications that have become partly irrelevant and hinder scientific progress. Perceiving the evolving structure of science is critical for NIH to make good decisions. Computational techniques are available now that produce “knowledge maps.” These maps can provide important information about the future of research.

Recommendation 8-2: The NIH should target individual NRSAs in emerging fields, interdisciplinary areas, and specific fields of interest. Such applications should be given priority in the awards process, and special review panels should be used as needed.

This approach will encourage scientists in the various fields to contribute to the task of identifying new areas. Individual awards can respond most rapidly to new initiatives. In addition they can be easily adjusted as fields mature or evolve in unanticipated directions. Moreover, a small number of awards can be very effective in attracting people to new fields and establishing standards. These awards should be made at both the predoctoral and postdoctoral levels.

The committee recognizes that such efforts are ongoing, but these efforts should be integrated across the institutes, including a formal structure to ensure a long-term vision.

Recommendation 8-3: Quantitative subject matter should be integrated into and required for training programs in all areas. Quantitative subjects include statistics, mathematics, physics, physical chemistry, computer science, and informatics.

The need for quantitative training is stressed throughout this report. With the overwhelming amount of new data becoming available, it is essential that scientists understand how to analyze and critically interpret the information. In all areas of biology and medicine, understanding biological processes and health issues on a quantitative basis will be of increasing importance. Although quantitative training is already prescribed in many cases, it is a necessity for all areas of biomedical, behavioral, and clinical research.

Footnotes

1

National Research Council. 2004a.

2

Rosenfield, P. L. 1992.