Remarks

Good afternoon, everyone! Thank you to Dr. Collins for his invitation to be this year's Lindberg-King Lecturer. I want to take a moment to recognize Donald Lindberg and Donald King, who worked together to contribute so much to the field of medical informatics, and to the National Library of Medicine. Thank you both for taking the time to be here today, and for your contributions to science and medical research.

This theme of cooperation is one that I want to focus on today. I have divided my talk into three parts: NSF and NIH's history and partnership; NSF's contributions to providing foundational knowledge key to biomedicine; and NSF's path forward on the frontiers of science and engineering.

First, I want to show a brief video about NSF and its unique role in our Nation's science research ecosystem.

NSF's founding mission is clear: "To promote the progress of science; to advance the national health, prosperity, and welfare; to secure the national defense." This responsibility to focus on basic, fundamental research across all fields of science and engineering complements NIH's concentration on biomedical science research that enhances the health of people worldwide. Our two agencies share similarities, including missions that grew out of comparable circumstances.

As you may know, NIH's early beginnings originated with the Marine Hospital Service in 1887, in a one-room laboratory founded by Joseph Kinyoun called the Laboratory of Hygiene. After the First World War, chemists wanted to set up an establishment in the private sector that would apply fundamental knowledge in chemistry to the lasting problems of medicine. However, after finding no success, supporters partnered with Louisiana Senator Joseph Ransdell to secure federal sponsorship, which established the present-day NIH in 1930.

NSF was formed in 1950, in response to a report by Vannevar Bush, the Director of the Office of Scientific Research and Development. Bush's report, "Science: The Endless Frontier," detailed how government could promote the best ideas from science and engineering research and education for greater social good, and was the catalyst for what is now NSF.

Every day, NSF and NIH prove the importance of federal funding in research. Our two agencies resulted from people who believed that applying resourceful thinking to longstanding societal problems would lead to the advancement of our nation. It is only natural that NSF and NIH would have a long, productive partnership.

For instance, since 2013, the joint NSF/NIH Smart and Connected Health (SCH) Program has accelerated the development and use of innovative approaches to support the transformation of healthcare from reactive and hospital-centered to preventative, proactive, evidence-based, and person-centered.

The program involves several NSF directorate and numerous NIH institutes; it includes investments in breakthrough ideas in a variety of areas and technologies.

We have also collaborated on the Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative, bringing the nation closer to a deeper understanding of nervous systems. We also partner on the Collaborative Research in Computational Neuroscience (CRCNS) solicitation with several international partners. This program provides a theoretical foundation and a rich set of technical approaches for understanding complex neurobiological systems, building on the theory, methods, and findings of computer science and neuroscience.

And NIH adopted the Innovation Corps program, or I-Corps, that was pioneered by NSF. I-Corps is a public-private network of scientists, engineers, business leaders, and entrepreneurs to strengthen our national innovation ecosystem. With NIH's participation, this effort is now providing entrepreneurial education to help biomedical researchers translate their innovations to the marketplace.

Pictured in the lower right of the screen is Stephen DiMagno, whose fundamental chemistry research led to new technology for therapeutic medicines. Among the first recipients of an NSF I-Corps grant, he founded a company that manufactures molecules that become imaging agents for managing cancer, cardiac disease, and neurological disorders.

These are just a few of the ways our NSF/NIH collaboration has contributed to scientific discovery. By working together to share data, ideas, and technologies we can enhance the impact of our investments and better address the grand challenges of our time.

Having talked about our shared history and partnership, let me turn next to NSF's contributions to foundational knowledge key to biomedicine.

When it comes to these grand challenges, NSF's motto "Where discoveries and discoverers begin" affirms our commitment to applying new thinking - and new thinkers - to unsolved mysteries. Our commitment to funding basic research across all fields has led to some significant medical breakthroughs. Here are a few examples of NSF-funded research with application to biomedicine.

For instance, NSF provided a substantial part of the basic research infrastructure that scientists drew upon to develop magnetic resonance imaging (MRI). From 1955 through the 1990s, NSF support for the underlying nuclear magnetic resonance (NMR) instrumentation amounted to $90 million, and in 1991 MRI pioneer and 2003 Nobel Prize winner Paul Lauterbur received NSF funding to launch the Center for Magnetic Resonance Technology for Basic Biological Research.

Our support translates discoveries revealed by basic research and applies them in specific ways to enhance society. Thirty-five years ago, NSF started the Small Business Innovation Research program to fund innovative 'start ups' with potential for growth. Today many federal agencies, including NIH, fund SBIR and its sister program in universities, the Small Business Technology Transfer.

NSF support through initiatives like SBIR has led to advances in fields such as synthetic biology, bio-scaffolding, and tissue engineering. This slide illustrates one example. The small business Miromatrix Medical developed technology to create bioengineered organs for human transplant.

In this example, a quadriplegic man experiences the sense of touch again. By connecting a robotic arm to a brain computer interface (BCI) implanted in his head, he is able to "feel" pressure on the robotic hand. The blueprint for the robotic arm came from NSF-funded basic research that looked at the neural activity of monkeys as they manipulated objects. This advancement is paving the way for touch-sensitive prosthetics.

Presently, our funding is supporting a diverse group of researchers at four U.S. universities plus research institutions in Japan and South Korea in researching soft robotics, specifically a polymer-based material that twists, moves, and senses, unlike traditional robots. The team is aiming to convert the material into artificial muscle, which could be a life-changing discovery for people with disabilities around the world.

The groundbreaking discoveries do not stop there. NSF-funded researchers studying how a bacterium's immune system fights off viruses uncovered a powerful gene-editing technique, which we all know as CRISPR-Cas9, which acts like a pair of molecular-sized scissors that researchers can wield to snip a segment of DNA. NSF funded the first grant for this research, and NIH took it from there. Today we look to CRISPR as a tool with the potential to fight genetic disorders like ALS and muscular dystrophy and maybe even prevent vector-borne infectious diseases.

Those are just a few of the examples of recent progress in scientific research, revealing the potential to change lives worldwide. And there are new frontiers to explore.

We've been looking at our history; now let's look to the present, and what lies ahead. First, I want to talk about an area that is important to both the NSF and NIH communities - data science - which will lead us directly into NSF's 10 Big Ideas for future investment. Combining computational advances and biomedical research can solve many enduring grand challenges, and data science may be the key to unlocking a transformational future of medicine.

The Big Data revolution is underway. With the rapidly increasing volume, variety, and velocity of data, innovative and fundamental research questions can now be addressed.

In addition to some of our data-focused partnerships I mentioned earlier, NIH and NSF also partner on the "Quantitative Approaches to Biomedical Big Data" Initiative, an example of NIH's commitment to utilizing computational science to reach the full potential of biomedical research. And NIH and NSF also co-chair an Interagency Working Group on Open Science.

Bioinformatics has substantially increased in importance to medicine. In the past, medicine has not been in a position to take advantage of bioinformatics, but more and more we can see the need and opportunity for collaboration between the data science and health research communities. Combining the science of medicine, the collection and mining of relevant data, and the intuition of the best doctors will revolutionize what medicine can accomplish.

Contributing to this promising area, NSF has made significant investments to build software and data infrastructure. We are advancing the capacity of our nation's academic researchers to capture, curate, and make accessible a growing body of increasingly heterogeneous life sciences data.

We invest in numerous data curation activities, such as the Protein Data Bank that is a partnership with NIH and DOE. As the need for data integration and scaling increases, new investments in data portals and aggregators are providing the ability for researchers to link broadly heterogeneous data sources.

NSF also invests in research into novel data technologies and methods that will bring enhanced capabilities to this infrastructure. We are funding research that is providing cutting-edge algorithms, data types, visualization approaches, and statistical methods. This research addresses the contemporary challenges that come with increases in the rate, volume, and complexity of data generated.

Data science is an important frontier - a field that needs to be further explored, and is ripe for future collaborations between our agencies.

So, we've seen examples of our investments and their impact, and the possibilities that are on the horizon. Where do we go from here?

NSF's commitment to advancing the progress of science has led us to identify 10 Big Ideas for Future Investment, shown in this slide.

These Big Ideas confront a range of compelling research challenges across all fields of science and engineering, from understanding the rules of life, to navigating the warming arctic, to leading the next quantum revolution, to examining the nature of dark matter and dark energy using novel particle, electromagnetic, and gravitational wave facilities. They also include big process ideas, which would change how we fund and conduct research. I'm going to talk about just four of these Big Ideas that are relevant to the theme of my talk: the impact of computation on biomedicine.

The first Big Idea that I want to highlight is Harnessing the Data Revolution.

I mentioned the data revolution earlier, and this Big Idea is NSF's vision for exploring that emerging frontier. The data deluge poses its own challenge - one that requires a bold, national-scale solution, with imaginative approaches to data science, to cyberinfrastructure, and to policy implications like balancing privacy and openness.

Given its history with founding NSFnet, a forerunner to today's internet, and its current portfolio of large facilities generating big data, NSF is well positioned to lead and partner with other agencies to execute this initiative. Our vision for harnessing big data can produce substantial benefits for fundamental research across all areas of science, from astrophysics to ecology to earth and atmospheric science to medicine. We envision turning translational data science into reality, allowing researchers to ask and answer fundamentally new questions and generate new knowledge. NSF will also build interdisciplinary teams to accelerate progress. A crucial part of harnessing big data means developing a comprehensive, research cyberinfrastructure ecosystem. It also means creating and nurturing a data-capable workforce.

This Big Idea brings together and greatly expands NSF's ongoing investments in data science and engineering. Working together with key partnerships with industry and other agencies will enable us to radically alter how we conduct scientific research in the future.

I want to focus for a minute on machine learning, which operates on big data. Machine learning is the universal connector that interweaves all of our Big Ideas; data science is changing the very nature of scientific inquiry, and machine learning has the potential to revolutionize all aspects of science. Because of its importance, NSF, NIH, and fellow science agencies like NIST, DOE and DoD have a special role in supporting machine learning, and Artificial Intelligence tools in general.

We are merely scratching the surface of our understanding of AI, yet the impact in medical research is already evident.

Advances in data analytics and machine learning are helping to accurately diagnose illnesses and personalize treatments, pushing us closer to an age where precision medicine is commonplace. Machine learning will allow physicians to efficiently convert significant amounts of data into knowledge, and then translate that knowledge into solutions.

For example, scientists are using the NSF-supported Blue Waters supercomputer to help understand HIV. One of the biggest barriers to creating truly effective therapies to combat the virus is that the exact structure of the HIV capsid has proven elusive - until just recently. Researchers Juan Perilla and the late Klaus Schulten at the University of Illinois successfully used the supercomputer to create a detailed molecular map of the HIV-1 capsid, as seen on this slide.

This is one example of the ways intelligent systems can change medicine. But algorithms are unlikely to completely replace humans; physicians and scientists will have to learn to work in collaboration with intelligent systems. And that leads me to the next Big Idea, Shaping the Human-Technology Frontier.

One of AI's most interesting lessons is that the most efficient way to utilize machines in addressing complex challenges may be to pair them with humans. Using machine learning to process the volumes of big data and latest information will aid in diagnostics and prognosis accuracy, and this will allow doctors to focus on quality time with patients. The belief is that AI will help make good doctors even better, influencing more accurate, informed decisions and more precise predictions.

The partnership is already proving to be beneficial. A recent study revealed that AI helped to significantly diminish the pathologist error rate in detecting metastatic breast cancer from lymph node biopsies. An AI-based computational system and a human pathologist were given images of lymph node cells, and asked to determine if the cells were cancerous. The AI-based approach had a 7.5 percent error rate, while a human pathologist had a 3.5 percent error rate. But when human and machine collaborated, the error rate was lowered to 0.5 percent, an 85 percent reduction in error.

That study is only scratching the surface of what humans and technology can collaborate to accomplish. Appearing on the screen is Suchi Saria. Dr. Saria and an NSF-funded team of researchers at Johns Hopkins University are using machine learning to analyze large volumes of medical records. Dr. Saria recently developed an AI program integrating data from the health records of more than 16,000 patients to identify 27 factors capable of predicting septic shock.

Dr. Saria's work is an example of what we can do when we invest in the future STEM workforce. One of NSF's priorities has been to make science more inclusive, because we don't know where the next groundbreaking discovery will come from. We want to increase participation in science and engineering from our brightest young minds, because our Ideas can’t reach full potential without talented, well-prepared scientists and engineers at all levels.

INCLUDES is the Big Idea that reflects NSF's belief that HOW we do research is just as important as WHAT we research, and we must work to strengthen and diversify the composition of the science and engineering workforce that we will depend on for future innovation.

NSF has long been a champion of initiatives to not only improve STEM education, but broaden participation among those traditionally underrepresented in STEM fields. For example, part of NSF's investment in bioinformatics research focuses on growing the next generation of data-aware scientists, through a broad spectrum of training programs that target undergraduate to mid-career scientists. But NSF also understands the importance of investing in instructional materials that prominently feature diverse role models that may inspire the scientists and physicians of the next generation.

The last Big Idea I want to mention is Convergence. Einstein once wrote, "To raise new questions, new possibilities, to regard old problems from a new angle requires creative imagination and marks real advance in science." This is the essence of convergence, a Big Idea that involves interdisciplinary teams coming together intentionally, in novel ways, to strategize a research plan that confronts challenges that know no disciplinary borders. In fact, the very notion of applying computation to biomedicine is one of different disciplines collaborating for the progress of science.

I want to show a brief video that illustrates the real-life impact of bringing together seemingly disparate disciplines for significant medical benefits.

Dr. Roth is one of the 223 Nobel Prize winners who have received NSF funding, and one of the 55 Nobel Laureates in the economic sciences that we have supported. His example of an economist revolutionizing the way kidney transplants are managed highlights the value of using new thinking to attack complex problems.

NSF has supported cross-disciplinary research for decades in our mission to strengthen the nation through advancing the progress of science. Just last month we announced $17.7 million in funding for 12 Transdisciplinary Research in Principles of Data Science (TRIPODS) projects. These TRIPOD projects are conducted at 14 institutions in 11 states, aiming to bring together the statistics, mathematics, and theoretical computer science communities to develop the foundations of data science.

There are other specific examples that I can point to that embody this concept of convergence. NSF recently awarded a large, long-term grant to UC San Francisco for a Center for Cellular Construction, which will be used to study and promote cellular engineering. The ultimate goal is to transform the field of cell biology into a quantitative discipline and use tools from engineering, the physical sciences, and computer science to create computerized machines out of living cells. The future of cellular engineering could lead to new disease-fighting strategies, and many more possibilities in areas we cannot imagine.

When convergence truly works, seemingly unrelated disciplines that may make unconventional partners come together to produce unexpected new applications. In the aftermath of the launch of the Hubble Space Telescope and discovering its blurred optics, astronomers and cancer researchers had a shared problem. They both needed to pinpoint critical patterns against a cluttered and often blurred background. Radiologists need to search images for microcalcifications as signs of breast cancer. These images are similar to images of the cosmos that astronomers study. Thanks to funding from an NSF grant, groups of astronomers and radiologists from Johns Hopkins, Georgetown, and the Space Telescope Institute were able to collaborate on new software that today allows radiologists greater ability to scrutinize mammograms for signs of breast cancer; and the field of digital mammography was born from this unusual partnership. As an astrophysicist, I appreciated that this technology also brought more clarity to the first-generation images from Hubble!

We have made substantial gains in science and health research, but in order to achieve even greater breakthroughs it is imperative that we invest in creative approaches like convergence. Having fresh ideas and original processes, and investing in state-of-the-art facilities, tools, and software will go a long way towards the health research breakthroughs that we are all eagerly seeking.

These partnerships - between humans and technology, between disciplines, between agencies - are what the future of innovation depends on. Together, we have supported a lot of high-risk studies that have resulted in indispensable contributions to society. However, our potential for discovery is so much higher. These emerging frontiers of science are ripe to be explored, and NSF believes that bringing people and disciplines together is the ideal approach to finding those transformative discoveries. Our challenge is to use the collaborations that will help make the breakthroughs of tomorrow a reality. We cannot shy from our commitment to promote the progress of science. We cannot slow our efforts to empower later generations to shape a brilliant future.

NSF and NIH's dedication to science is the root that has kept this Nation established as a global leader in science and technology, and tomorrow's scientists and physicians will count on our discoveries to cross the next frontiers of science. Our dedication to discovery is our connection with the future, in the spirit of the old proverb, "A society grows great when old men plant trees whose shade they know they shall never sit in."

The results of our shared commitment to basic research has brought us to a time of unprecedented capacity for innovation. If we take the opportunity we have in front of us to move science forward by deepening our partnerships and exploring new ways of doing science and imaginative ways to it, our history - a history of incredible accomplishments - foreshadows a frontier of endless possibilities.

Thank you!