[ad_1]
Generative AI technologies can write poetry and computer programs or create images of teddy bears and cartoon character videos that looks like something out of a Hollywood movie.
Now, new artificial intelligence technology is generating blueprints for microscopic biological mechanisms that can edit your DNA, pointing to a future in which scientists will be able to combat diseases with greater precision and speed than today.
Described in a research paper published on monday By a Berkeley, California startup called Profluent, the technology is based on the same methods that power ChatGPT, the online chatbot that launched the AI boom after its launch in 2022. The company is expected to present the paper next month at the annual meeting of the American Society for Gene and Cell Therapy.
Just as ChatGPT learns to generate language by analyzing Wikipedia articles, books, and chat logs, Profluent’s technology creates new gene editors after analyzing huge amounts of biological data, including microscopic mechanisms that scientists already use to edit human DNA.
These gene editors are based on Nobel Prize-winning methods involving biological mechanisms called CRISPR. CRISPR-based technology is already changing the way scientists study and fight disease and illnessproviding a way to alter genes that cause inherited diseases, such as sickle cell anemia and blindness.
Previously, CRISPR methods used mechanisms found in nature: biological material obtained from bacteria that allows these microscopic organisms to fight germs.
“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. “The system has learned from nature to create them, but they are new.”
The hope is that the technology will eventually produce gene editors that are nimbler and more powerful than those that have been perfected over billions of years of evolution.
On Monday, Profluent also said it had used one of these AI-generated gene editors to edit human DNA and that it was “open source” this editor, called OpenCRISPR-1. That means it allows individuals, academic labs and companies to experiment with the technology for free.
AI researchers often Open source the underlying software that powers your AI systems, because it allows others to take advantage of your work and accelerate the development of new technologies. But it is less common for biological laboratories and pharmaceutical companies to open source inventions like OpenCRISPR-1.
Although Profluent offers open access to the gene editors generated by its AI technology, it does not offer open access to the AI technology itself.
The project is part of a broader effort to develop artificial intelligence technologies that can improve healthcare. Scientists at the University of Washington, for example, are using the methods behind chatbots like OpenAI’s ChatGPT and image generators like Midjourney to create completely new proteins (the microscopic molecules that power all human life) as they work to accelerate the development of new vaccines and drugs.
(The New York Times has defendant OpenAI and its partner, Microsoft, over copyright infringement claims involving artificial intelligence systems that generate text).
Generative AI technologies are driven by what scientists call a neural network, a mathematical system that learns skills by analyzing large amounts of data. The Midjourney image maker, for example, is based on a neural network that has analyzed millions of digital images and the captions that describe each of those images. The system learned to recognize links between images and words. So when you ask him for a picture of a rhino jumping off the Golden Gate Bridge, he knows what to do.
Profluent’s technology is powered by a similar AI model that learns from sequences of amino acids and nucleic acids, the chemical compounds that define the microscopic biological mechanisms that scientists use to edit genes. Basically, it analyzes the behavior of CRISPR gene editors taken from nature and learns how to generate completely new gene editors.
“These AI models learn from sequences, whether they are sequences of characters, words, computer codes or amino acids,” said Profluent CEO Ali Madani, a researcher who previously worked in software giant Salesforce’s AI lab.
Profluent has not yet put these synthetic gene editors through clinical trials, so it is unclear whether they can match or surpass the performance of CRISPR. But this proof of concept shows that AI models can produce something capable of editing the human genome.
Still, it is unlikely to affect healthcare in the short term. Fyodor Urnov, a gene-editing pioneer and scientific director of the Innovative Genomics Institute at the University of California, Berkeley, said scientists had no shortage of natural gene editors they could use to combat disease and illness. The bottleneck, he said, is the cost of forcing these publishers to conduct preclinical studies, such as safety, manufacturing and regulatory reviews, before they can be used in patients.
But generative AI systems often have enormous potential because they tend to improve rapidly as they learn from increasing amounts of data. If technology like Profluent’s continues to improve, it could eventually allow scientists to edit genes in much more precise ways. The hope, Dr. Urnov said, is that this can, in the long term, lead to a world where medications and treatments are quickly tailored to each person, even faster than we can today.
“I dream of a world where we have CRISPR on demand in a few weeks,” he said.
Scientists have long warned against using CRISPR for human enhancement because it is a relatively new technology that could have unwanted side effects, such as causing cancer, and they have warned against unethical uses, such as genetically modifying human embryos.
This is also a concern with synthetic gene editors. But scientists already have access to everything they need to edit embryos.
“A bad actor, someone who is unethical, doesn’t care whether or not they use an AI-created editor,” Dr. Fraser said. “They’ll just go ahead and use what’s available.”
