NSF invests in new approaches to drive the future of biotechnology
The U.S. National Science Foundation announced seven new awards totaling more than $10 million through the NSF Molecular Foundations for Biotechnology (NSF MFB) program, supporting collaborative research that uses machine learning methods to advance fundamental research in biomolecular systems with the potential to spur new biotechnologies. These new awards build upon four made as part of the program in 2021, bringing the total NSF investment to more than $15 million.
Machine learning, a form of artificial intelligence that allows a system to learn from experience, has driven recent innovations in many areas and holds the potential to tremendously accelerate scientific progress and transform the nature of research in the physical and life sciences. In recent years, machine learning was used to predict the structure of a protein based on its sequence, furthering researchers' ability to characterize, understand and harness the power of a critical building block of life. The new awards will seek to create similar breakthroughs in areas, including protein design, RNA-based regulation and protein-DNA assemblies, paving the way for advances in biotechnology, medicine and industry that are critical to the nation's long-term economic and national security.
"NSF's MFB program aligns with critical agency goals to provide the nation with the capacity to approach more complex questions than ever before," said NSF Assistant Director for Mathematical and Physical Sciences Sean L. Jones. "NSF's investment will bring foundational advances in biotechnology that will impact health care, manufacturing, sustainability and more."
The application of machine learning technologies to biomolecular questions is the most recent focus area of NSF's multiyear effort to support fundamentally new approaches in molecular sciences and drive new directions in biotechnology. The focus area in 2021 was novel chemical biology tools to drive innovations in biotechnology.
"By bringing together teams that can cross these new computing and analytic methods with expertise from chemical and biological sciences, the research has the potential to yield findings that can transform our country," said Margaret Martonosi, NSF assistant director for Computer and Information Science and Engineering.
"Biology and the millions of ways life has adapted and evolved hold extraordinary promise to address critical issues from the need for sustainable sources of energy to the call for new and more efficient medicines and vaccines," said NSF Assistant Director for Biological Sciences Joanne Tornow. "By supporting the use of advancements in other fields of science and engineering to answer long-standing questions in biology, the MFB program is bringing us closer to delivering on that promise."
The Fiscal Year 2022 awards made by the program, and the lead institutions, are:
- MFB: Targeting the dark proteome by machine-learning-guided protein design (Rutgers University-New Brunswick).
- MFB: Integrating deep learning and high-throughput experimentation to rapidly navigate protein fitness landscapes for non-native enzyme catalysis (University of Wisconsin-Madison, Morgridge Institute for Research, Indiana University).
- MFB: Deciphering RNA-based regulatory logic with interpretable machine learning (New York University, Cornell University).
- MFB: NSF-BSF: Data-Adaptive and Metamorphosis Machine Learning Architectures for Generative Protein Design of Metal Biosensors (University of Kansas, University of Haifa).
- MFB: Accelerating the Discovery of Novel Liposome Formations with Origins-of-Life Insights, Laboratory Automation, and Machine Learning (Fordham University).
- MFB: Novel Graph Neural Networks to Understand, Predict, and Design Allosteric Transcription Factors (Georgia Tech).
- MFB: Deep-Learning Enabled Structure Prediction and Design of Protein-DNA Assemblies (University of Washington).