Recommendations of The Biomedical Information Science and Technology Initiative Implementation Group
April 2000
The fundamental objective put forward in the report of the Working Group on Biomedical Computing of the Advisory Committee to the Director of the National Institutes of Health on the Biomedical Information Science and Technology Initiative (BISTI Report) is the following:
"To make optimal use of information technology, biomedical researchers need, first of all, the expertise to marry information technology to biology in a productive way. New hardware and software will be needed, together with deepened support and collaboration from experts in allied fields. Inevitably, those needs will grow as biology moves increasingly from a bench-based to a computer-based science, as models replace some experiments and complement others, as lone researchers are supplemented by interdisciplinary teams. The overarching need is for an intellectual fusion of biomedicine and information technology."
To support this objective, the report identified four primary recommendations. The BISTI Implementation Group acknowledges the wisdom of the recommendations put forward in the report and has developed a series of implementation recommendations tied to the recommendations of the BISTI Report. The four primary recommendations of the BISTI Report are listed below together with the implementation recommendations of the BISTI Implementation Group.
BISTI Report Recommendation #1:
The NIH should establish between five and twenty National Programs of Excellence in Biomedical Computing devoted to all facets of this emerging discipline, from the basic research to the tools to do the work. It is the expectation that those National Programs will play a major role in educating biomedical-computation researchers.
Implementation Recommendations:
As noted in the report, the establishment of National Programs of Excellence in Biomedical Computing (NPEBC) would create in the community foci of expertise in biomedical information science and technology that would speed research in this area and provide an infrastructure to enable the training for a future generation of cross-disciplinary scientists. The report recommended the establishment of "between five and twenty National Programs of Excellence in Biomedical Computing devoted to all facets of this emerging discipline". It is not entirely clear at this point that all potentially qualified institutions are ready to submit coherent programs that would qualify as ready for a National Program of Excellence designation. As noted in the report (page 6) "Biomedical computing is not a recognized discipline, and despite the extraordinary demand for people with good education in both biomedicine and computing, only a few cross-disciplinary training programs exist." It is clear that there is a range of problems that might potentially benefit from the establishment of a NPEBC. In the report alone there are seven very broad examples provided. At this time there is not a clear consensus on which areas would stand to benefit the most from the near term establishment of a NPEBC.
Pre-NPEBC
For these reasons it is proposed to proceed with a two-pronged approach towards the establishment of NPEBC. The first approach would focus on the establishment of planning awards (P20) that would lead to eligibility for competitive NPEBC (see appended draft solicitation). The planning grants should provide support and incentive for qualified teams to be brought together, and the pre-NPEBC selection and maturation process should help to reveal the most useful programs and scientific areas to target for larger investments. The pre-NPEBC solicitation would be issued as a trans-NIH initiative to ensure that the full range of problems that might benefit from support can be targeted. As currently envisioned the grants could provide extensive flexibility in size and time frames up to three years to match the potentially disparate needs of different institutions. The awards would provide support for multidisciplinary teams to come together and plan for the structure of a National Program of Excellence, the establishment of core facilities, interactive activities, and the creation of a training and education environment to support biomedical computing. In addition these planning grants would support the initiation or expansion of a limited number of highly interactive research projects that span the interface of biomedical research and biomedical information science and technology.
NPEBC
Individual or groups of institutes or centers would issue solicitations for NPEBC. The solicitations would be focussed around particular areas of opportunity for partnership in the areas of biomedical computing and biomedical research, and would be issued in parallel with, or subsequent to, the release of the NIH-wide solicitation of planning grants. The timing of the release of NPEBC solicitations would be determined by the institutes or centers involved and the maturity of the bioinformatics and computational research in the targeted scientific area. It is anticipated that in some areas solicitations for NPEBC could be successful without prior investments in pre-NPEBC, and therefore NPEBC solicitations for some areas might be issued in parallel with the solicitation of pre-NPEBC.
- As recommended in the BISTI report NPEBC are intended to support interdisciplinary teams that target:
- Promoting bio-informatics and bio-computational research in the service of advanced biomedical research
- Developing useful and interoperable informatic and computational tools for biomedical research
- Establishing mutually beneficial collaborations between biomedical researchers and informatics and computation researchers
- Training a new generation of bio-informatics and bio-computation scientists who are truly cross-disciplinary, i.e. have training in biomedical science and one or more traditionally unrelated fields, including but not limited to informatics or computer science, or electronics or electrical engineering, theoretical or applied mathematics, physics etc.
BISTI Report Recommendation #2:
To make the growing body of biological data available in a form suitable for study and use, the NIH should establish a new program directed toward the principles and practice of information storage, curation, analysis, and retrieval (ISCAR).
Implementation Recommendations:
The report correctly highlights the rapidly expanding collections of data that require storage, curation, analysis, and retrieval, as well as the critical need for technological developments that will enhance database interoperability. The solution put forward is to create a program that will support the collection of information as well as the development of new tools for storage, management, and access. This area is clearly one of critical importance, but also is of a sufficient breadth and magnitude that an open-ended program might not yield maximum progress in any one area. The recommendation of the BISTI Implementation Group is that the greatest progress is likely to come from targeting specific areas of need or scientific opportunity that could benefit from focused ISCAR efforts and promoting research in the principles and practice of ISCAR in the context of targeted institute objectives. Focussed ISCAR programs linked to priority biomedical research programs could yield useful interoperable databases, as well as result in fundamental insights in to improved principles and practices for ISCAR and enhanced biomedical research discoveries for the targeted area.
It is proposed that each year NIH institutes select major scientific initiatives from within or across institutes that would benefit from, and drive research into, the principles and practice of ISCAR. As a portfolio of ISCAR investments develops the NIH could provide information on that portfolio through a web accessible NIH catalogue of these efforts. The NIH could also facilitate the dialogue between various ISCAR efforts and promote the development of interoperability, where appropriate, between NIH funded ISCAR efforts.
BISTI Report Recommendation #3:
The NIH should provide additional resources and incentives for basic research (through R01 grants) to provide adequate support for those who are inventing, refining, and applying the tools of biomedical computing.
Implementation Recommendations:
As noted in the report, additional resources and incentives for basic research should be provided to establish and enhance the field of biomedical computing. The report specifically recommends the use of the R01 mechanism. While the R01 mechanism offers great flexibility, the utility of this mechanism in the current system has been limited for research in computing science and technology. Research in biomedical computing is frequently cost intensive, requiring significant hardware investments to support the final project. These projects also frequently have associated risk as minimal preliminary information may exist that the idea or approach will be productive. In addition, these projects are not necessarily hypothesis driven. The potential high cost balanced by the risk and frequent lack of preliminary validation would argue for a phased mechanism for support. Given the speed with which scientific research is progressing and new technologies emerging, the need for rapid discovery and/or development is also clear and would argue against the conventional approach to a staged process that would require submission, selection, and award of an application, together with completion of concept validation before a competitive review of a request for expanded project support can be considered.
For this reason it is proposed that both broad and more specific solicitations for biomedical computing research could utilize the recently developed R21/R33 mechanism in a Phased Innovation in Biocomputing Award. The new R21/R33 phased mechanism has been developed by the National Cancer Institute to support the rapid testing and development of innovative new technologies and has met with tremendous enthusiasm from the research community. Specific features of the use of this mechanism for the Phased Innovation in Biocomputing Award would include:
- Single submission and evaluation of both the R21 and R33 as one application.
- Expedited transition of feasibility phase to an expanded research or development phase.
- Flexible budgets.
- Flexible staging of feasibility and research or development phases.
Small business is playing an increasing role in the development of new tools for biomedical computing. The NIH should clearly capture the mutual value of the activities in small business in this area. The BISTI Implementation Group is proposing to have a parallel solicitation using the SBIR and STTR mechanisms to encourage small business participation. The SBIR/STTR solicitation would be structured to have comparable cost and duration requirements to the Phased Innovation in Biocomputing Award, and have common review panels and review criteria to encourage productive use of the SBIR set aside of funds.
The first issuance is recommended to be a pair of broad "open window" solicitations for all relevant areas of biomedical information science and technology (see appended draft soliciations). One solicitation would support applications through the R21/R33 mechanism, and the other solicitation would use the SBIR and STTR mechanisms for small business. These solicitations would be in parallel with or followed by more specific solicitations in areas of identified need.
BISTI Report Recommendation #4:
The NIH should foster a scalable national computer infrastructure. To assure that biomedical researchers can gain access to the computing resources they need beyond their desktops, the NIH should provide financial resources to increase computing capacity, both local and remote. The purpose of this recommendation is to establish a balanced infrastructure for all computational needs.
The BISTI Implementation Group realizes the importance of working together with other agencies to explore the role NIH should play in supporting enhanced computing infrastructure to support biomedical computing. Discussions have begun between the Director, NIH and the NSF. The Implementation Group has no specific recommendations other than to continue such discussions with NSF and any other relevant agencies, as well as to consider the potential need for supporting infrastructure through the support to the NPEBC and pre-NPEBC.
Coordination across NIH Institutes, Centers and Programs
In order to maximize our opportunities to benefit from advances in biomedical information science and technology, and to learn from our experiences in investing in biocomputing, a NIH BISTI Consortium should be established. The consortium could be modeled after, but distinct from, the very successful Bioengineering Consortium at the NIH. The Consortium would be charged with facilitating the sharing of information across the NIH on emerging scientific opportunities in biocomputing, working together in developing initiatives in biocomputing, addressing issues that arise in managing biocomputing initiatives, and developing strategies such as BISTI related meetings or symposia, to enhance the value of our investments in biomedical information science and technology.
The BISTI Implementation Group recommends that NIH IC Directors identify an appropriate representative of their institute/center to participate in a BISTI Consortium. A first task of the Consortium could be to complete and shepherd the issuance of the recommended solicitations. The Implementation Group also recommends that coordination of this consortium and trans-NIH BISTI activities should reside in the OD of the NIH. It is suggested that this coordination might reside in the Office that will oversee NIH Bioengineering and Bioimaging activities. It is also the opinion of the Implementation Group that BISTI be considered with the same level of priority and visibility as bioengineering and bioimaging and that inclusion of this scientific area in to the Office should be taken in to consideration in the selection of the Director for the coordinating office. In addition to coordinating the BISTI Consortium, this Office would also be tasked with representing the NIH on BISTI related issues to the Administration and Congress, as well as a focal point for coordination with other agencies on computing infrastructure issues.
