Researchers claim to perform gene editing with AI, offer model as CRISPR alternative

Artificial Intelligence is now creating detailed plans for tiny biological processes capable of editing your DNA, paving the way for a future where scientists can combat illnesses and diseases with remarkable precision and speed.

A research, outlined in a paper published on Monday by Profluent, a start-up in Berkeley, California, uses the same techniques powering ChatGPT, the chatbot that sparked the AI revolution upon its debut in 2022.

Profluent is set to showcase their findings next month at the annual meeting of the American Society of Gene and Cell Therapy.

"CRISPR-based gene editors derived from microbes, while powerful, often show significant functional tradeoffs when ported into non-native environments, such as human cells.

"Artificial intelligence (AI) enabled design provides a powerful alternative with potential to bypass evolutionary constraints and generate editors with optimal properties," the paper read.

Just as ChatGPT learns to generate language by studying lots of text, Profluent's tech makes new gene editors by studying huge amounts of biological info, including tiny processes scientists use to edit human DNA. Or so it claims.

These gene editors rely on CRISPR which is a gene-editing technology used to modify DNA sequences.

Gene editing involves modifying the genetic code within an organism to change or enhance certain characteristics or functions. It offers a method for modifying genes responsible for hereditary conditions like sickle cell anaemia and blindness.

The methodology involved the use of large language models (LLMs) trained on biological diversity. A dataset comprising over one million CRISPR operons was curated by mining 26 terabases of assembled genomes and meta-genomes systematically.

The models were then deployed to generate a significantly increased number of protein clusters across CRISPR-Cas families found in nature, and to tailor single-guide RNA sequences for Cas9-like effector proteins.

The efficacy and specificity of the generated gene editors were evaluated, with several showing comparable or improved activity relative to SpCas9, the prototypical gene editing effector, despite being 400 mutations away in sequence.

Profluent's technology is powered by an AI model that learns from sequences of amino acids and nucleic acids. These are the chemical compounds that make up the tiny biological processes scientists use to edit genes. Basically, it studies the actions of CRISPR gene editors found in nature and figures out how to create brand-new ones.

"These AI models learn from sequences — whether those are sequences of characters or words or computer code or amino acids," said Profluent’s chief executive, Ali Madani, a researcher who previously worked in the AI lab at the software giant Salesforce.

Generative AI technologies operate through a neural network, which is a mathematical system that acquires skills by studying large datasets, according to scientists.

"They have never existed on Earth," said James Fraser, a professor and chair of the department of bioengineering and therapeutic sciences at the University of California, San Francisco, who has read Profluent’s research paper. "The system has learned from nature to create them, but they are new."

While Profluent is making the gene editors created by its AI technology, named OpenCRISPR-1, open source, they're not sharing the AI technology itself.

Profluent hasn't tested these artificial gene editors in clinical trials yet, so it's uncertain if they can match or surpass the effectiveness of CRISPR.

However, the trial run demonstrates that A.I. models can create tools capable of editing the human genome, it claims.

Image source: Unsplash