Team science exemplified

Selection of the independent researchers and range of expertise

The large-scale collaborative project consisted of participating investigators who have successful basic research programs in the relevant areas of research.  The participating investigators were selected because each is a successful NIH and particularly NIGMS-funded investigator in basic research both in the field of injury research but more than half of the group had successful federally funded research programs in a very broad distribution of disciplines and fields.  These included many diverse disciplines including clinical medicine (surgery, internal medicine, anesthesiology, infectious disease, endocrinology, health policy, nuclear medicine, pathology, pharmacology, physiology, critical care medicine, and emergency medicine), animal science, biochemistry, biophysics, microbiology, immunology, molecular and cellular biology, physical chemistry and biochemistry, engineering and material sciences, genomics and genetics, computational biology, bioinformatics, and biostatistics.  Each investigator had extensive experience with the genomic portion of the field.  However, few had experience with the global, systematic approach necessitated by whole genome bioinformatics and few investigators had worked on a collaborative study that required the integration of such sophisticated genomic and proteomic analyses in the clinical setting.

Establishing and fostering collaboration within and among the groups

The program’s work began by organizing into individual core groups of specific disciplines and fields (genomics, proteomics, clinical trials, computational biology, informatics, model systems, administrative).   Each group was led by a core director, who facilitated the team interactions to identify goals, responsibilities,and an infrastructure that would support the individual and programmatic milestones.

• The core groups were initially faced the requirements to develop robust protocols to generate high quality samples of sufficient quantity for genomics and proteomics analyses that could be performed in an intensive care setting or mouse laboratory.  The challenge was to develop protocols that balanced high quality science with technologies and in the clinical trials group, that were feasible when performed by trained research nurses and technicians in critically ill patients.  This required two years to test the feasibility and practicality of using these advanced technologies in the critical care setting.

• The patient-oriented research (clinical trials) core was charged to develop an infrastructure that could provide consistent patient care and generate quality samples among the various clinical sites that had research personnel of diverse backgrounds and training.  The trauma and burns clinical groups were the first groups to adopt a consensus-driven approach to define the critically ill trauma patient and create clinical treatment protocols (standard operating procedures, SOPs) and this approach was adopted by the analytical, data, and computational groups.

• The information dissemination and data collection core (IDDC) was challenged to create many data collection and organizational tools for the collection, validation, and long-term storage of the data.  This required the creation of several essential software tools, including a laboratory information system to organize the blood and tissue samples, a web-accessible, relational database to capture the demographic, clinical, and outcomes data, a data warehouse combining all of the clinical, genomics, and proteomics data, and web sites to share information.

• The computational analysis and modeling group was charged to provide statistical and computational input as experiments and protocols were actually being designed to insure that robust and meaningful data sets would be generated.   The latter challenge went on to become a critical issue not only for this program, but also for the other Glue Grant-funded projects – all of which had underestimated the bioinformatics and computational challenges that laid ahead, not only in terms of cost, but also in the lack of current available tools to analyze and derive biological meaning from the amassing, sophisticated data sets.

Teamwork – the collaborative approach of the groups

A huge psychosocial challenge to team science is the optimization of modes of communication within groups and between groups.  The research program developed a cohesive and functional scientific, analytic, and administrative infrastructure that had continuously evolved throughout the first three years of funding.  Program activities during Year 1 could be viewed as communication and progress within the individual groups.  Although the program took a proactive approach in internal and external communications, program leaders found that keeping communication channels open and functional continued to be a challenge (especially for the administration core) and required a tremendous input of time and effort from all core groups. Program leaders employed a number of passive and interactive methods of communication within the program to help the groups work effectively together. The program’s internal communication strategies and approaches included quarterly face-to-face meetings (nearly all in Chicago, the “half-way point” for most participating sites), site visits to inter-group laboratories and facilities, email discussions, faxes, one-to-one telephone calls, group conference calls, and the program’s internal web site for the participating investigators.

Year 2 activities encompassed more broad communication and progress amongst the various groups to develop and test consensus-driven standards and protocols that would ultimately drive the overall research program. The truly best work of the individual groups and the participating investigators as a whole was accomplished at the the face-to-face meetings of the investigators who met in the morning as individual groups and came together as one, large group in the late afternoon and evening every three months for ten years.  Interestingly, the investigators were reluctant to adopt video-conferencing technology for the program and cited cameras, microphones, audio playback, special software, and high bandwidth network cabling requirements as detracting features of videoconferencing.

Difficulty decisions for growth and sustainability

In interesting feature of the program was that initial budgets for major core activities were put together for the original 2001 grant application without program infrastructure already in place or a complete and accurate knowledge of what it would take in dollar costs to carry out all of the proposed activities.  In consideration of the overall program, the program leaders recognized early in Year 4 that the group needed to carry out internal assessments of the important components and use this information to evolve the program in Year 5, so that the program could continue seamlessly into Year 6 if renewed for another five years. As is inevitable with any large-scale program, some of the initially proposed objectives proved to be either unfeasible or unaffordable by the steering or advisory committees.

The principal investigator and steering committee with guidance from the advisory committee recognized that the Year 4 priorities should be focused on the following key elements: (1) the human studies; (2) quantitative and high throughput analyses; and (3) on-going data analysis and interpretation. The Glue Grant program had demonstrated that it could accomplish what it promised and in 2006 was renewed for another five years.  This, however, was not accomplished without significant reorganization and a few difficult decisions. As a result of its team science experience during the first four years, the team collected and analyzed data, fully recognized cost/benefit constraints, faced analytical and financial constraints, brainstormed with the participating investigators and consortium members, and followed the advice of its advisory committee.  No matter the size of the group and the funding available, hard decisions about what to fund and what to do remain an important challenge.  The final decisions regarding these tradeoffs must be made in the best interests of the program and its goals as a whole.

Infrastructure applicable for future studies

Within a newly expanded data interpretation core funded starting in Year 6, the program created clinical data interpretation groups (clinical DIGs) as the approach for the large-scale and sophisticated data interpretation tasks that the program would face in the new funding period.  Focused multidisciplinary project teams composed of biostatisticians, systems biologists, basic science investigators knowledgeable about pathways, and clinicians, among others, participate in joint evaluation of data in real time to “discover” the data and uncover the biological significance of the results.  The first clinical DIG meeting took place during the quarterly participating investigators meeting in January 2006 and these clinical DIGs continue today.  The program found that this structure is vital to derive real biological knowledge from the massive, complex, and integrated data sets.

The program’s overall strategy is to sustain the infrastructure, resources, tools, and methodologies developed in the Glue Grant through funding collaborations with other federal funding agencies, industry, and organizations.  One of the major strengths of the program is the group of individuals involved along with their collegiality and professional attributes.  The program has fostered a high level of cooperation, mutual respect, and collegiality among the community of investigators than previously existed.  This form of research is not attractive to all independent investigators and as much as one third of those who began with the program lost interest and did not follow through with the program throughout its entire existence.  Of those who did follow through, they gained a tremendous advantage to their own education and learning.  Many returned quarterly to report their progress and to learn the results from others who are leaders in their own disparate fields.  This is a very effective form of adult education for already well-established, well-educated, professors and deans of science in leading academic medical centers and universities here in the US.

Contact

Ronald Tompkins, MD, ScD	Lyle Moldawer, PhD	Ronald Maier, MD	Wenzhong Xiao, PhD
617-726-3447	352-265-0494	206-744-3299	617-724-7261

Visit the Glue Grant program website

See Less