The NIH Bioinformatics and Computational Biology Roadmap

1
The NIH Bioinformatics and Computational Biology
Roadmap

• Peter Lyster PhD
• Program Director, Computational Biology and
  Bioinformatics
• National Institute for Biomedical Imaging and
  Bioengineering (NIBIB) at the National Institutes
  of Health (NIH)
• For the Coalition for Academic Scientific
  Computation (CASC) Winter meeting, February 4,
  2004,Washington DC

2
User oriented mission statement

• In ten years, we want every person involved in
  the biomedical enterprise---basic researcher,
  clinical researcher, practitioner, student,
  teacher, policy maker---to have at their
  fingertips through their keyboard instant access
  to all the data sources, analysis tools, modeling
  tools, visualization tools, and interpretative
  materials necessary to do their jobs with no
  inefficiencies in computation or information
  technology being a rate-limiting step.

3
Computational Biology at the NIHwhy, whence,
what, whither

• WhyBecause computation and information
  technology is an invaluable tool for
  understanding biological complexity, which is at
  the heart of advance in biomedical knowledge and
  medical practice.
• You cant translate what you dont
  understand---Elias Zerhouni, Director of the
  National Institutes of Health, commenting on the
  relationship between basic research and
  translational research, that transforms the
  results of basic research into a foundation for
  clinical research and medical practice.

4
Computational Biology at the NIHwhy, whence,
what, whither

• WhenceComputation and information technology
  were originally used as add-ons, to add value to
  experimental and observational results that had
  sufficiently simple patterns that they could be
  discerned by observation. Often the computing
  technology was an almost invisible partner to the
  experiments. For example, the 1951
  Hodgkin-Huxley Nobel Prize work that elucidated
  the bases of electrical excitability included
  calculations that were done on an
  electromechanical calculator, and would not have
  been feasible by hand or slide ruleyet it is not
  often cited as an example of the importance of
  calculating technology.

5
Computational Biology at the NIHwhy, whence,
what, whither

• WhatToday computation is at the heart of all
  leading edge biomedical science. For leading
  examples, consider this past years Nobel prizes
• Structure of voltage-gated channelsrequired
  sophisticated computation for image
  reconstruction for x-ray diffraction data, the
  mathematical techniques for which were the
  subject of a previous Nobel prize.
• Discovery of water channelsThe experimental work
  required augmentation by bioinformatics for
  identification of water channel genes by sequence
  homology.
• Magnetic resonance imagingA large share of the
  prize work was for the mathematical and
  computational techniques for inferring structure
  and image from NMR spectra.

6
Computational Biology at the NIHwhy, whence,
what, whither

• Institutes and Centers at NIH support substantial
  development and implementation of computation and
  information technology embedded in biomedical
  research.
• Informatics is a key component of the NIH Roadmap
  Initiative

7
Roadmap Activities Computation

• New Pathways to Discovery
• National Centers for Biomedical Computing
• Building Blocks, Biological Pathways, and
  Networks
• Re-engineering the Clinical Research Enterprise
• National Electronic Clinical Trials and Research
  Network (NECTAR)
• Dynamic Assessment of Patient-Reported Chronic
  Disease Outcomes
• Trans NIH Informatics Committee (TNIC)

8
Present State of Computational Biology Practice

• Essentially all leading-edge biomedical research
  utilizes significant computing.
• Development and initial implementation of methods
  are largely the product of collaborations with
  overlapping expertise---biologists who have
  substantial expertise in computing with computer
  scientists and other quantitative scientists who
  have substantial knowledge of biology. Computer
  scientists and other quantitative scientists with
  little knowledge of biology are generally unable
  to contribute to the development of biomedical
  computing tools.

9
The Paradox of Computational Biology--Its
successes are the flip side of its deficiencies.

• The success of computational biology is shown by
  the fact that computation has become integral and
  critical to modern biomedical research.
• Because computation is integral to biomedical
  research, its deficiencies have become
  significant rate limiting factors in the rate of
  progress of biomedical research.

10
Some important problems with biomedical computing
tools are

• They are difficult to use.
• They are fragile.
• They lack interoperability of different
  components
• They suffer limitations on dissemination
• They often work in one program/one function mode
  as opposed to being part of an integrated
  computational environment.
• There are not sufficient personnel to meet the
  needs for creating better biological computing
  tools and user environments.

11
Computational Biology at the NIHwhy, whence,
what, whither

• WhitherThe NIH Bioinformatics and Computational
  Biology Roadmap
• Was submitted to NIH Director Dr. Elias Zerhouni
  on May 28, 2003
• Is the outline of an 8-10 year plan to create an
  excellent biomedical computing environment for
  the nation.
• Has as its explicit most ambitious goal Deploy a
  rigorous biomedical computing environment to
  analyze, model, understand, and predict dynamic
  and complex biomedical systems across scales and
  to integrate data and knowledge at all levels of
  organization.

12
1-3 year roadmap goals relatively low difficulty

• 1. Develop vocabularies, ontologies, and data
  schema for defined domains and develop prototype
  databases based on those vocabularies,
  ontologies, and data schema
• 2. Require that NIH-supported software
  development be open source.
• 3. Require that data generated in NIH-supported
  projects be shared in a timely way.
• 4. Create a high-prestige grant award to
  encourage research in biomedical computing.
• 5. Provide support for innovative curriculum
  development in biomedical computing
• 6. Support workshops to test different methods or
  algorithms to analyze the same data or solve the
  same problem.
• 7. Identify existing best practice/gold standard
  bioinformatics and computational biology products
  and projects that should be sustained and
  enhanced.
• 8. Enhance training opportunities in
  bioinformatics and biomedical computing.

13
1-3 year roadmap goals moderate difficulty

• 1. Support Center infrastructure grants that
  include key building blocks of the ultimate
  biomedical computing environment, such as
  integration of data and models across domains,
  scalability, algorithm development and
  enhancement, incorporation of best software
  engineering practices, usability for biology
  researchers and educators, and integration of
  data, simulations, and validation.
• 2. Develop biomedical computing as a discipline
  at academic institutions.
• 3. Develop methods by which NIH sets priorities
  and funding options for supporting and
  maintaining databases.
• 4. Develop a prototype high-throughput global
  search and analysis system that integrates
  genomic and other biomedical databases.

14
4-7 year roadmap goals relatively low difficulty

• 1. Supplement existing national or regional
  high-performance computing facilities to enable
  biomedical researchers to make optimal use of
  them.
• 2. Develop and make accessible databases based on
  domain-specific vocabularies, ontologies, and
  data schema.
• 3. Harden, build user interfaces for, and deploy
  on the national grid, high-throughput global
  search and analysis systems integrating genomic
  and other biomedical databases.

15
4-7 year roadmap goals moderate difficulty

• 1. Develop robust computational tools and methods
  for interoperation between biomedical databases
  and tools across platforms and for collection,
  modeling, and analyzing of data, and for
  distributing models, data, and other information.
• 2. Rebuild languages and representations (such as
  Systems Biology Markup Language) for higher level
  function.

16
4-7 year roadmap goals high difficulty

• 1. Ensure productive use of GRID computing
  through participation of biologists to shape the
  development of the GRID.
• 2. Develop user-friendly software for biologists
  to benefit from appropriate applications that
  utilize the GRID.
• 3. Integrate key building blocks into a framework
  for the ultimate biomedical computing
  environment.

17
8-10 year roadmap goals relatively low difficulty

• 1. Employ the skills of a new generation of
  multi-disciplinary biomedical computing
  scientists

18
8-10 year roadmap goals moderate difficulty

• Produce and disseminate professional-grade,
  state-of-the art, interoperable informatics and
  computational tools to biomedical communities. As
  a corollary, provide extensive training and
  feedback opportunities in the use of the tools to
  the members of those communities.

19
8-10 year roadmap goals high difficulty

• Deploy a rigorous biomedical computing
  environment to analyze, model, understand, and
  predict dynamic and complex biomedical systems
  across scales and to integrate data and knowledge
  at all levels of organization.

20
Initial Steps on the Roadmap Plan I

• We have released a funding announcement, and
  received proposals, for the creation of four NIH
  National Centers for Biomedical Computing. Each
  Center is to serve as the node of activity for
  developing, curating, disseminating, and
  providing relevant training for, computational
  tools and user environments in an area of
  biomedical computing. We hope ultimately to
  establish eight centers.

21
Initial Steps on the Roadmap Plan II

• We are preparing a funding announcement for
  investigator-initiated grants to collaborate with
  the National Centers. Instead of having big
  science and small science compete with each
  other, we will create an environment in which
  they will work hand in hand for the benefit of
  all science.

22
Initial Steps on the Roadmap Plan III

• We are preparing a funding announcement for work
  on creating and disseminating curricular
  materials that will embed the learning and use of
  quantitative tools in undergraduate biology
  education for future biomedical researchers. We
  are committed to pressing a reform movement in
  undergraduate biology education to ensure an
  adequate number of quantitatively trained and
  able biomedical researchers in the future.

23
Initial Steps on the Roadmap Plan IV

• We are in the initial stages of establishing a
  formal assessment and evaluation process. A
  possible form is that an external group of
  scientists will establish criteria by which to
  evaluate the program, and a professional survey
  research group will work with the scientists to
  implement the ongoing assessment and evaluation
  plan, so that prompt and appropriate mid-course
  corrections and tuning will take place.

24
Key Features of the NIH Bioinformatics and
Computational Biology Roadmap Process

• Every component goes through NIH peer review
  system.
• Larger components are by cooperative agreement
  rather than grant, with active continued
  participation by NIH program staff.
• There is complete transparency about the rules
  and the process (except for the confidentiality
  necessary for peer review).
• Assessment and Evaluation are built in from the
  start.
• Program, review, and evaluation are independent
  of each other.

25
SOFTWARE DISSEMINATION REQUIREMENTS FOR NIH
NATIONAL CENTERS FOR BIOMEDICAL COMPUTING. (As
expressed in the funding announcement for this
project) There is no prescribed single license
for software produced in this project. However
NIH does have goals for software dissemination,
and reviewers will be instructed to evaluate the
dissemination plan relative to these goals 1)
The software should be freely available to
biomedical researchers and educators in the
non-profit sector, such as institutions of
education, research institutes, and government
laboratories. 2) The terms of software
availability should permit the commercialization
of enhanced or customized versions of the
software, or incorporation of the software or
pieces of it into other software packages. 3) The
terms of software availability should include the
ability of researchers outside the center and its
collaborating projects to modify the source code
and to share modifications with other colleagues
as well as with the center. A center should take
responsibility for creating the original and
subsequent "official" versions of a piece of
software, and should provide a plan to manage the
dissemination or adoption of improvements or
customizations of that software by others. This
plan should include a method to distribute other
user's contributions such as extensions,
compatible modules, or plug-ins. The application
should include written statements from the
officials of the applicant institutions
responsible for intellectual property issues, to
the effect that the institution supports and
agrees to abide by the software dissemination
plans put forth in the proposal.
26
Possible areas of productive interaction with
other agencies

• with DOE on microbial science and nanoscience and
  biotechnology
• with DARPA on microbial science and on
  nanoscience and biotechnology
• with USDA on nutrition and agricultural science
• with NIST on data and software standards and on
  nanoscience
• with NSF on biology at all levels, on integrating
  biomedical computational science with the
  cyberinfrastructure initiative, on fostering
  interdisciplinary collaborative science, on
  nanoscience, and on biology education
• f. with NASA and NOAA on environmental issues
  related to health

27
National Institute for Biomedical Imaging and
Bioengineering (NIBIB)

• Dr. Roderic Pettigrew Director
• Improve health through fundamental
  discoveries, design and development, and
  translation and assessment of technological
  capabilities in biomedical imaging and
  bioengineering, enabled by relevant areas of
  information science, physics, chemistry,
  mathematics, materials science, and computer
  sciences.

28
NIBIB Computation Activities

• Biomedical Information Science and Technology
  Consortium (BISTIC)
• Neuroinformatics
• Human Brain Project (HBP)
• Collaborative Research in Computational
  Neuroscience (CRCNS)
• Neuroimaging Informatics Technology Initiative
  (NIfTI)
• Interagency Modeling and Analysis Group (IMAG)

29
Interagency Modeling and Analysis Group (IMAG)

• Formed in 2003, lead by NIBIB
• Purpose To promote modeling and analysis
  methods in biomedical systems
• Interagency initiative on Multiscale Modeling

30
Interagency Modeling and Analysis Group (IMAG)

• Participants
• 13 NIH components (NIBIB, NIDA, NIGMS, NINDS,
  NCI, NIMH, NHGRI, NCRR, NICHD/NCMRR, NLM, NIEHS,
  OD, and CSR)
• 3 NSF directorates (ENG, CISE, and BIO)
• DOD
• DARPA
• TATRC
• NASA

31
NIBIB Program Areas

• Mathematical Models and Computational Algorithms
• Bioinformatics and Medical Informatics
• Human-Computer Interface, Image Displays,
  Perception, and Image Processing
• Imaging Device Development
• Imaging Agent and Molecular Probe Development
• Tissue Engineering

• Biomaterials
• Medical Devices and Implant Science
• Therapeutic Agent Delivery Systems and Devices
• Biosensors
• Biomechanics and Rehabilitation Engineering
• Platform Technologies
• Nanotechnology
• Remote Diagnosis and Therapy
• Surgical Tools and Techniques

32
Mathematical Models and Computational Algorithms

• Multiscale, structural and functional modeling
• New, novel modeling methodology (nonlinear and
  systems methods)
• Clinical decision algorithms
• Statistical methods and data reduction models
• Imaging algorithms (distortion correction and
  motion detection)
• Data analysis methods
• Tangible molecular models

33
Bioinformatics

• Data acquisition, management and processing
• Data mining and data analysis
• Networked tools for transfer of images and
  radiological reports (GRID)
• Digital atlases, gene expression maps,
  probabilistic maps
• Knowledge-based reporting systems
• Mapping and visualization of function and
  diseases (genotype and phenotype)
• Medical informatics
• Biostatistics

34
Image Processing

• Segmentation and registration
• Image analysis, pattern recognition,
  computer-aided diagnosis
• Multi-modal imaging analysis (PET, MR,)
• Neuroimaging
• Mammography
• Perceptual modeling
• Dosage Radiography

35
Remote Diagnosis and Therapy

• Remote-monitoring and quantification of images
• Data and model integration for critical care
• Wearable sensors and data fusion
• Haptics and tele-diagnostics
• Neurophysiological interoperative monitoring
• Internet-based home healthcare
• Remote-management of disease

36
Surgical Tools and Techniques

• Computer-assisted surgery (Haptics)
• Simulated surgical training
• Image-guided interventions

37
Future Directions at NIBIB

• Interagency Modeling and Analysis Group (IMAG)
• Systems Biology/Tissue Engineering
• Imaging Informatics
• Data Integration
• Large-scale Databases

38
NIBIB Program Contactshttpwww.nibib.nih.gov

• Modeling / Bioinformatics / Neuroprosthesis /
  Telehealth Technologies / Biomechanics /
  Rehabilitation
• Grace Peng penggr_at_mail.nih.gov
• Biosensors / Tissue Engineering
• Chris Kelley - kelleyc_at_mail.nih.gov
• Biomaterials / Nanotechnology
• Peter Moy - moype_at_mail.nih.gov
• Bioinformatics / Imaging Informatics
• Peter Lyster - lysterp_at_mail.nih.gov
• Imaging
• John Haller - hallerj_at_mail.nih.gov
• Alan McLaughlin mclaugal_at_mail.nih.gov
• Yantian Zhang yzhang1_at_mail.nih.gov
• Training
• Meredith Temple-OConnor - templem_at_mail.nih.gov