Profluent Achieves Human Genome Editing Milestone Using OpenCRISPR-1: The First AI-Generated, Open-Source Gene Editor

Profluent, a Berkeley-based company specializing in AI-driven protein design, recently announced the OpenCRISPR-1^TM project, presenting the global debut of the first ever open-source, AI-developed gene editor. This introduction signifies a significant achievement in the precise editing of the human genome using customizable gene editors made entirely with AI technology.

Transforming Gene Editing: The AI Revolution in CRISPR Technology

Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) technology, first developed over ten years ago, facilitates the precise modification of DNA sequences in living organisms. Its potential applications range from treatments for genetic disorders through to investigating disease mechanisms. Here, unlike the earlier gene editing methods such as CRISPR-Cas9, OpenCRISPR-1 generates molecules that are entirely synthetic and do not occur in nature.

Functionally, OpenCRISPR-1, as an AI-generated gene editor, comprises a Cas9-like protein and guide RNA, developed entirely with Profluent’s large language models (LLMs). During OpenCRISPR’s training process, Profluent’s AI used and learned from huge amounts of sequence data and biological contexts to produce millions of diverse synthetic CRISPR-like proteins. This process exponentially broadened the scope of all known CRISPR families. Essentially, Profluent’s technology uses extensive biological data to create new gene editors, similar to how ChatGPT learns language from various sources like Wikipedia and conversations.

Ali Madani, Profluent’s CEO, conveyed, “These A.I. models learn from sequences — whether those are sequences of characters or words or computer code or amino acids.”

Available for Public Use: Profluent’s AI-Crafted OpenCRISPR-1

In a bid to democratize the technology, Profluent launched OpenCRISPR-1TM as an initial open-source release. This move allows the AI-designed gene editor to be freely licensed for ethical research and commercial purposes in the development of CRISPR-based medicines. The decision to open-source OpenCRISPR-1 marks a pivotal moment in gene editing, and Profluent believes that this initiative will accelerate discovery and innovation in the field by leveraging AI to design tailored medications precisely suited to various needs. AI’s distinctive capability lies in its capacity to redefine and construct gene editing systems entirely anew, a feat unattainable through traditional protein engineering approaches. By broadening the types of CRISPR tools available and providing foundational components for further development, AI is expected to make gene editing therapies more accessible and affordable.

With a vision to expand the accessibility of CRISPR therapies, the company plans to collaborate with leading research institutions and pharmaceutical developers across the drug development spectrum, ultimately striving to make CRISPR medicines available to a wider range of patients and for a broader spectrum of disorders.

“Attempting to edit human DNA with an AI-designed biological system was a scientific moonshot,” commented Ali Madani, who serves as Profluent’s co-founder and Chief Executive Officer. “Our success points to a future where AI precisely designs what is needed to create a range of bespoke cures for disease. To spur innovation and democratization in gene editing, with the goal of pulling this future forward, we are open-sourcing the products of this initiative.”

The Role of AI in Advancing Gene Editing

At the core of this achievement lies the use and integration of AI, as the company utilized LLMs trained on extensive sequence and biological data. Profluent’s team curated a database containing 5.1 million Cas9-like proteins, which served as the foundation for training the AI model to generate potential proteins for CRISPR applications. Leveraging examples found in nature, the AI model was able to construct novel gene editors from scratch, culminating in the identification of OpenCRISPR-1. This protein exhibits comparable performance to Cas9 but with significantly reduced off-target effects, ensuring greater precision and minimal DNA damage.

A sweeping trend in healthcare is the integration of gene editing technologies like SHERLOCK and DETECTR, facilitating swift identification of infectious diseases such as COVID-19. Simultaneously, companies like Atomwise, Deep Genomics, and Valo are integrating gene editing into their drug discovery endeavors, reshaping the landscape of treatment development. Beyond gene editing, AI is fueling innovations across various healthcare domains, from bone marrow analysis software to drug discovery platforms, aiding in matching cancer patients with the most suitable treatment drugs, and even generating novel drugs.

Through using AI to study microscopic biological processes involved in DNA editing, these types of advanced technology aim to transform the fight against illnesses and diseases with exceptional accuracy and effectiveness. Profluent has published the scientific details of the protein’s development in a preprint publication and presented this development at the 27th Annual Meeting of the American Society of Gene and Cell Therapy (ASGCT) held in Baltimore, with an oral session focusing on this novel AI-driven method for tailor-making gene editors.

Author

Bernice Lottering