Generative AI technology can create poems, computer programs, or videos of images of teddy bears or cartoon characters like those in Hollywood movies.
Now, new AI technologies are generating blueprints for microscopic biological mechanisms that can edit DNA, pointing to a future in which scientists can fight disease and disease with greater precision and speed than they do today. I am.
The technology, as described in a research paper published Monday by Berkeley, California startup Profluent, is based on the same technique that powers ChatGPT, the online chatbot that sparked an AI boom after its release in 2022. Based on. The paper will be presented at the annual meeting of the American Society for Gene and Cell Therapy next month.
Just as ChatGPT learns to produce language by analyzing Wikipedia articles, books, and chat logs, Profluent's technology involves subtle mechanisms that scientists already use to edit human DNA. , create a new gene editor after analyzing vast amounts of biological data.
These gene editors are based on a Nobel Prize-winning technique that involves a biological mechanism called CRISPR. CRISPR-based technology is already changing the way scientists study and fight disease and disease, providing a way to modify genes that cause inherited diseases such as sickle cell anemia and blindness.
Previous CRISPR techniques used mechanisms found in nature: biological substances collected from bacteria that allow microorganisms to fight bacteria.
“They have never existed on Earth,” said James Fraser, professor and chair of the Department of Bioengineering and Therapeutic Sciences at the University of California, San Francisco, who read the Profluent research paper. “Systems have been learning from nature to create them, but they are new.”
The hope is that this technology will eventually create gene editors that are more agile and powerful than those honed over billions of years of evolution.
Profluent also announced Monday that it is using one of its AI-generated gene editors to edit human DNA and is “open sourcing” the editor, called OpenCRISPR-1. That means individuals, academic institutions, and companies will be able to experiment with the technology for free.
AI researchers often open source the underlying software that powers their AI systems. This is so that other researchers can build on their work to accelerate the development of new technologies. But it's less common for biology labs and pharmaceutical companies to open source inventions like his OpenCRISPR-1.
Profluent is open sourcing the gene editor generated by its AI technology, but not the AI technology itself.
This project is part of a broader effort to build AI technology that can improve healthcare. For example, scientists at the University of Washington are using the techniques behind chatbots like OpenAI's ChatGPT and image generators like Midjourney to create entirely new proteins, the tiny molecules that power all human life. and are accelerating the development of new vaccines and vaccines. medicine.
(The New York Times sued OpenAI and its partner Microsoft for copyright infringement claims involving artificial intelligence systems that generate text.)
Generative AI technology is powered by what scientists call neural networks, mathematical systems that learn skills by analyzing vast amounts of data. For example, image creator Midjourney is powered by a neural network that has analyzed millions of digital images and the captions that describe each of those images. The system has learned to recognize connections between images and words. So when Golden Gate asks for an image of a rhino jumping off a bridge, he knows what to do.
Profluent's technology is powered by similar AI models that learn from sequences of amino acids and nucleic acids, the compounds that define the microscopic biological mechanisms that scientists use to edit genes. Essentially, you will learn how to analyze the behavior of CRISPR gene editors extracted from nature and generate entirely new gene editors.
“These AI models learn from sequences, including letters, words, computer code, and sequences of amino acids,” said Profluent CEO Ari, a researcher who previously worked at software giant Salesforce's AI lab. Madani said.
Profluent has not yet conducted clinical trials with these synthetic gene editors, so it is unclear whether they will match or exceed the performance of CRISPR. But this proof of concept shows that an AI model can produce something that can edit the human genome.
Still, it is unlikely to have an impact on health care in the short term. Fyodor Urnov, a pioneer in gene editing and scientific director of the Institute for Innovative Genomics at the University of California, Berkeley, says scientists have no shortage of natural gene editors that can be used to fight disease and disease. said. The bottleneck, he said, is the cost of requiring editors to conduct preclinical studies, including safety, manufacturing and regulatory reviews, before using them on patients.
But generative AI systems often have great potential because they tend to improve rapidly as they learn from increasingly large amounts of data. If technologies like Profluent continue to improve, they could eventually allow scientists to edit genes in a much more precise way. In the long term, Dr. Urnov said, he hopes this could lead to a world in which personalized medicines and treatments can be delivered quickly and even more quickly than is possible today.
“I dream of a world where CRISPR will be available on demand within weeks,” he said.
Scientists have long warned against using CRISPR for human enhancement, as it is a relatively new technology that can cause undesirable side effects, such as inducing cancer, and has long warned against using it for human enhancement, such as genetically modifying human embryos. have warned against the unethical use of
This is also a concern for synthetic gene editors. But scientists already have access to everything they need to edit embryos.
“Bad actors and unethical people don't care whether they use an AI-generated editor or not,” Dr. Fraser said. “They just take advantage of what's available.”