Generative AI enters CRISPR gene editing field

Generative AI techniques can write poetry and computer programs, or create images of teddy bears and videos that look like cartoon characters from Hollywood movies.

Now, new artificial intelligence technology is generating blueprints for microscopic biological mechanisms that can edit DNA, promising a future in which scientists can fight disease with greater precision and speed than today.

The technology, described in a research paper published Monday by a Berkeley, Calif., startup called Profluence, is based on the same approach as ChatGPT, the online chatbot that sparked the AI ​​revolution after its release in 2022. upsurge. The company expects to present the paper next month at the annual meeting of the American Society for Gene and Cell Therapy.

Just as ChatGPT learned to generate language by analyzing Wikipedia articles, books, and chat logs, Proflus technology creates new gene editors after analyzing large amounts of biological data, including the microscopic mechanisms that scientists already use to edit human DNA.

These gene editors are based on a Nobel Prize-winning approach involving a biological mechanism called CRISPR. CRISPR-based technology has changed the way scientists study and fight disease, providing a way to alter genes that cause genetic disorders such as sickle cell anemia and blindness.

Previously, CRISPR methods used mechanisms discovered in natural biological materials collected from bacteria to enable these microorganisms to fight off bacteria.

James Fraser, professor and chair of the Department of Bioengineering and Therapeutic Sciences at the University of California, San Francisco, who has read the Profluences research paper, said they have never existed on Earth. The system learns from nature to create them, but they are new.

The hope is that the technology will eventually lead to gene editors that are more flexible and powerful than those honed over billions of years of evolution.

On Monday, Profluence also said it has used one of its artificial intelligence-generated gene editors to edit human DNA and is open sourcing the editor, called OpenCRISPR-1. This means it allows individuals, academic labs and companies to experiment with the technology for free.

AI researchers often open source the underlying software that drives their AI systems because it allows others to build on their work and accelerates the development of new technologies. But it’s not common for biolabs and pharmaceutical companies to open source inventions like OpenCRISPR-1.

Although Profluu has open sourced the gene editor generated by its AI technology, it has not open sourced the AI ​​technology itself.

The project is part of a broader effort to build artificial intelligence technology that could improve health care. For example, scientists at the University of Washington are using the methods behind chatbots like OpenAIs ChatGPT and image generators like Midjourney to create entirely new proteins, the microscopic molecules that drive all human life, while working to accelerate the development of new vaccines and drugs.

(The New York Times has sued OpenAI and its partner Microsoft, accusing them of copyright infringement of the artificial intelligence system that generated the text.)

Generative AI technology is driven by what scientists call neural networks, mathematical systems that learn skills by analyzing large amounts of data. For example, the image creator Midjourney is based on a neural network that analyzes millions of digital images and the captions that describe each image. The system learned to recognize connections between images and text. So when you ask it for an image of a rhinoceros jumping off the Golden Gate Bridge, it knows what to do.

Profluences technology is driven by a similar artificial intelligence model that learns compounds from amino acid and nucleic acid sequences that define the microscopic biological mechanisms scientists use to edit genes. Essentially, it analyzes the behavior of CRISPR gene editors from nature and learns how to generate entirely new gene editors.

Profluences CEO Ali Madani, who previously worked in the artificial intelligence lab of software giant Salesforce, said these artificial intelligence models learn from sequences, whether those sequences are sequences of characters, single words, computer code sequences or amino acid sequences.

Profluence has not yet put these synthetic gene editors into clinical trials, so it’s unclear whether they will be able to match or exceed CRISPR’s performance. But this is proof of concept that artificial intelligence models can produce something capable of editing the human genome.

Still, it’s unlikely to impact health care in the short term. Scientists have no shortage of naturally occurring gene editors they could use to fight disease, said Fyodor Urnov, a pioneer in gene editing and scientific director of the Institute for Innovative Genomics at the University of California, Berkeley. The bottleneck, he said, is the cost of pushing these editors through preclinical studies (such as safety, manufacturing and regulatory reviews) before they can be used in patients.

But generative AI systems often have huge potential because they tend to improve quickly as they learn from more and more data. If technology like Profluences continues to improve, it may eventually allow scientists to edit genes in a more precise way. Dr. Urnov said the hope is that in the long term, this could lead to a world where drugs and treatments can be tailored to individuals faster than we can today.

“I dream of a world where we have CRISPR available on demand in a matter of weeks,” he said.

Scientists have long warned against using CRISPR to enhance human functions because it is a relatively new technology that can have unwanted side effects, such as causing cancer, and warned against unethical uses, such as genetically modifying human embryos. .

This is also a concern for synthetic gene editors. But scientists already have access to everything they need to edit embryos.

Dr. Fraser said a bad actor, an unscrupulous person, is not concerned about whether they use an editor created by artificial intelligence. They just keep going and use what’s available.