2025 Phil & Sandra Nudelman Endowed Lecture: The AI Frontier β Shaping the Future of Drug Design
[HPP] David BakerMarch 25, 20251h 19min
25 connectionsΒ·40 entities in this videoβAdvancements in AI-Driven Protein Design
- π‘ Dr. David Baker, a Nobel laureate, is at the forefront of computational protein design, creating entirely new proteins from scratch.
- π― AI is accelerating drug design and delivery, reshaping understanding of diseases, and enabling the creation of novel therapeutics.
- π§ The process reverses biological information flow, starting with a desired function to design a protein, its amino acid sequence, and then a DNA sequence.
- π RFdiffusion, a deep learning generative AI method inspired by image generation, is used to design completely new protein structures by progressively denoising random inputs.
Medical Applications of De Novo Proteins
- π Designed proteins can bind snake venom components, offering a potential alternative to traditional antivenoms.
- π₯ New proteins target inflammation pathways, showing greater effectiveness than existing drugs like Enbrel in animal models for autoimmune diseases.
- π AI-designed immunomodulators for cancer, such as IL21 receptor mimics, activate the immune system more effectively against pancreatic cancer.
- π¦ Antibody-like structures (nanobodies and full antibodies) can be designed to bind viruses like influenza, potentially simplifying pharmaceutical manufacturing.
- π§ Proteins designed to sandwich amyloid beta and tau peptides effectively suppress amyloid formation, providing tools for neurodegenerative disease research.
Technological Innovations with Designed Proteins
- π¬ Transmembrane nanopores with customizable sizes can be designed for new types of sensors, including those that integrate with silicon nitride chips.
- 𧬠Proteins can be designed to bind specific DNA sequences, offering tools for gene targeting and repair applications.
- π Nanomaterials like icosahedral structures have led to a potent COVID vaccine and are being developed as advanced delivery vehicles.
- βοΈ Proteins with multiple stable states can be engineered to change shape in response to effectors, enabling precise control over biological signaling.
Contributions to Sustainability
- π§ͺ De novo enzyme design allows for the creation of highly proficient catalysts for specific chemical reactions, approaching the activity of native enzymes.
- βοΈ Small proteins are designed to mimic photosynthesis, positioning chlorophyll molecules to build larger light-harvesting systems.
- β»οΈ Research is underway to design enzymes capable of breaking down plastics and other polymers, with positive results on microplastics.
- π Proteins can template inorganic mineral growth, such as zinc oxide, opening avenues for new semiconductor materials.
Future Outlook and Challenges
- β οΈ Biosecurity is a key concern, with calls for DNA synthesis logging to track potential misuse of powerful design methods.
- π While designed proteins show promise, scalability for mass production and a deeper understanding of human biology remain significant bottlenecks.
- π€ The success of protein design relies heavily on interdisciplinary collaboration and the creation of high-quality, accurate data sets for AI training.
- π‘ The field is moving towards personalized medicine, with the potential to design highly customized therapeutics for individual patient profiles.
Knowledge graph40 entities Β· 25 connections
How they connect
An interactive map of every person, idea, and reference from this conversation. Hover to trace connections, click to explore.
Hover Β· drag to explore
40 entities
Chapters19 moments
Key Moments
Transcript289 segments
Full Transcript
Topics15 themes
Whatβs Discussed
Artificial Intelligence (AI)Drug DesignProtein DesignDe Novo Protein DesignRFdiffusionGenerative AIProtein StructuresNanoporesDNA Binding ProteinsNanomaterialsEnzyme DesignPhotosynthesisPlastic DegradationBiosecurityPersonalized Medicine
Smart Objects40 Β· 25 links
PeopleΒ· 5
ConceptsΒ· 24
CompaniesΒ· 3
MediasΒ· 3
ProductsΒ· 5