Juan LiAI4Science
Reading updates
EN中

AI-generated weekly synthesis · selected by a reproducible rule, not curated by a human · not individually human-reviewed · verify original sources before citing

Protein Design’s AI Revolution: Inside David Baker’s “Communal Brain”

What changed

“I have this idea of a communal brain.” David Baker, PhD, told me as I sat in his office at the University of Washington (UW) surrounded by colorful and complex figurines of protein structures. It was the one-year anniversary of his Nobel Prize in Chemistry win. Just outside his doors, a lab of more than 100 researchers was united by the shared ambition to design proteins from scratch (or de novo) for powerful applications across pharmaceuticals, vaccines, biosensors, and more. This “communal brain” housed at the UW Institute for Protein Design (IPD), where Baker led as director, was hard at work developing deep learning methods that could achieve atomic precision. A small protein composed of 100 amino acids had an astronomical 20¹⁰⁰ possible sequences. Yet, only a vanishingly tiny fraction could fold into stable, functional structures. Misplacing a residue by an angstrom could mean the difference between a drug binding tightly to its target or complete failure. For the antibody drug market worth hundreds of billions of dollars, Nathaniel Bennett, PhD, former postdoctoral researcher in the Baker lab, says AI-guided antibody design that bypasses the need for time-consuming experimental screens has long been a “holy grail” for a breadth of indications, including cancer and autoimmune disease. Last November, Bennett and colleagues published a Nature paper demonstrating that full length de novo antibodies could bind user-specified epitopes. AI models could now construct antibody loops, the key region involved in binding that has been historically challenging to design due to its flexible nature. Despite this technological leap, AI-designed proteins that were manufacturable, remained stable in the body, and avoided unwanted side effects, were still a step away. The gap fueled an industry debate over whether generating de novo medicines was even possible. When I asked Baker to separate the hype from reality, he didn’t hesitate. “The reality is that we can now design proteins on a computer,” Baker explained in our video interview. “The hype is that for therapeutics, there’s a lot more than the basic activity of a protein binding or catalyzing a reaction. Whether de novo proteins will revolutionize medicine will require improving our understanding of the biology.” Nobel guests Bennett is continuing molecular design research as a co-founder at Xaira Therapeutics. Th…

Sources and evidence

外部发现 · David Baker · Original source ↗ · Site entry ·

Conjecture Machines: AI agents and the new validation bottleneck in science

What changed

Conjecture Machines: AI agents and the new validation bottleneck in science Over the past year, agents have transformed what it means to code. Scientific research may be next — from proposing novel hypotheses, to designing experiments, to discovering algorithms that improve on the best that humans have designed. This raises urgent questions for policymakers and science funders, the biggest being how to validate the coming wave of AI-generated ideas. New approaches are also needed to ensure that scientists can access agents, that datasets are agent-ready, and that peer review is kept afloat. We sat down with 10 researchers and engineers from Google DeepMind to try to figure out what comes next. Microbiologist José Penadés and his team at Imperial College London took most of a decade to work out how a family of superbugs spreads antibiotic resistance. The result was unpublished, known only inside his lab. Then, in 2024, he described the problem to Co-Scientist, an AI agent from Google DeepMind. Within two days, Co-Scientist returned five potential explanations, ranked in priority. Number 1 was the same hypothesis Penadés’s team had spent so long working to prove: that some superbugs acquire tails from viruses and use them as “keys” to jump between host species. Stunned, his first thought was that his computer had been compromised, so he emailed Google to check. Confirmed: no peeks taken. If he could have gone back in time, that insight would have saved his team years. An AI agent such as Co-Scientist is a large language model (LLM)- based system given a goal and the tools to pursue it. Unlike a query-answering chatbot, an agent can plan how to achieve the goal you give it, breaking it down into steps, running multiple subagents and processes in parallel, and detecting and correcting errors as it goes. It can also engage with the wider world, for example by calling online databases and tools, or writing code to operate robots. The arrival of AI agents is timely. Researchers face a rapidly growing “burden of knowledge”, with more to learn before they can meaningfully contribute. The questions that count in drug discovery, climate modelling, materials design, and biology are becoming too complex and interdisciplinary for human teams to tackle at pace. Science depends in part on a social infrastructure that has gradually evolved: labs, teams, institutions, peer r…

Sources and evidence

外部发现 · Google DeepMind · Original source ↗ · Site entry ·