Nurix Therapeutics Presents Data at the AACR 2025 Annual Meeting Highlighting the Transformative Potential of Its Proprietary DEL-AI Platform Leveraging Machine Learning to Speed the Discovery of Novel Drugs

Rhea-AI Summary

Nurix Therapeutics unveiled groundbreaking data at the AACR 2025 Annual Meeting showcasing their innovative DEL-AI platform, which combines DNA-encoded libraries with machine learning to accelerate drug discovery.

The company's DEL Foundation Model, developed in collaboration with Loka and supported by Amazon Web Services, demonstrates remarkable capabilities in:

• Accurately predicting novel drug binders for therapeutic targets
• Working with as little as 50% amino acid sequence similarity
• Identifying potential treatments for previously "undruggable" targets
• Inferring binders from chemical space outside training data

This AI-powered platform represents a significant advancement in drug discovery, potentially streamlining the identification of small molecule drugs, protein degraders, and molecular glues. The technology leverages Nurix's extensive database of over five billion unique DEL compounds screened against hundreds of disease targets and E3 ligase proteins.

The presentation, titled "DEL-AI: Proteome-wide in silico screening of multi-billion compound libraries using machine learning foundation models," validates the platform's ability to perform virtual DEL experiments and accurately predict experimental results.

Positive

• Development of first-in-class DEL Foundation Model capable of predicting novel drug binders, potentially accelerating drug discovery process
• Model requires only 50% amino acid sequence similarity to predict binders, expanding potential drug target range
• Platform can identify binders for previously 'undruggable' targets, opening new therapeutic possibilities
• Model demonstrates ability to predict binders from chemical space outside training data, showing versatility
• Strategic collaboration with Loka and AWS provides enterprise-grade infrastructure for the platform

Negative

• Platform is still in early development stage with no proven commercial success yet
• Technology requires significant computational resources and infrastructure investment
• No specific metrics provided on time or cost savings compared to traditional drug discovery methods

Nurix’s DEL-AI platform uses a first-in-class DEL Foundation Model trained on the Company’s proprietary DNA encoded library data

Nurix’s DEL Foundation Model can accurately predict novel binders to therapeutically relevant targets, including many targets considered undruggable, with the potential to accelerate the discovery of novel drugs

SAN FRANCISCO, April 28, 2025 (GLOBE NEWSWIRE) -- Nurix Therapeutics, Inc. (Nasdaq: NRIX), a clinical-stage biopharmaceutical company focused on the discovery, development and commercialization of targeted protein degradation medicines, today presented data that demonstrate the potential of its DEL Foundation Model to enable the rapid in silico identification of novel binders for a broad range of therapeutically relevant proteins, addressing a key barrier in the discovery and development of small molecule drugs. These results were presented at the American Association of Cancer Research (AACR) Annual Meeting in Chicago, IL, which is being held from April 25-30, 2025.

“Nurix’s DEL-AI platform has the potential to accelerate the discovery of breakthrough small molecule drugs—whether they be protein degraders, molecular glues, or inhibitors—by enabling ready-access to tractable chemical matter for an expansive set of proteins, especially those previously considered beyond the reach of drug discovery organizations,” said Gwenn M. Hansen, Ph.D., chief scientific officer of Nurix. “Our team has leveraged the rich datasets generated from rigorously controlled screenings of our customize collection of over five billion unique DEL compounds against hundreds of disease targets and E3 ligase proteins to construct a powerful suite of machine learning models and tools. By directly integrating the sampling density provided by DEL compound repertoires with primary protein sequence information, our model can learn a generalizable structure activity relationship capable of predicting novel binders for nearly any disease-relevant protein target.”

“Our DEL-AI engine is a potential game changer, allowing us to substantially accelerate drug discovery workflows and efficiently identify therapeutic candidates for our wholly owned pipeline and our current and future discovery partnerships,” said Arthur T. Sands, M.D., Ph.D., president and chief executive officer of Nurix. “This powerful research engine is a result of our significant expertise and strategic investments in DEL methodology and our machine learning platform.”

Nurix’s presentation at the AACR 2025 Annual Meeting, titled: “DEL-AI: Proteome-wide in silico screening of multi-billion compound libraries using machine learning foundation models,” described the development of a first-in-class foundation model that was trained on the Company’s high quality, proprietary DEL data. Nurix’s DEL Foundation Model is able to perform virtual DEL experiments on prospective protein target sequences to accurately predict novel binders to a large proportion of therapeutically relevant targets, including many targets considered undruggable. In plots of virtually predicted vs. experimentally-derived DEL screens against therapeutically relevant proteins, Nurix’s DEL Foundation Model demonstrated the ability to accurately predict the experimental results, including experimentally validated binders. Success of the DEL Foundation Model was found to correlate to the degree of similarity of query sequences to proteins within the DEL training set, and data demonstrated that the current model requires as little as 50% amino acid sequence similarity of a query protein to training data to enable binder prediction. Nurix’s model was also capable of inferring binders from chemical space not represented in the training set, suggesting that the model is capable of both protein sequence and chemical structure generalizations.

The development of the DEL Foundation Model was led by Nurix in collaboration with Loka, a Silicon Valley-based software development firm, and supported by Amazon Web Services (AWS), leveraging AWS SageMaker and AWS managed MLflow to provide enterprise-grade reliability and scalable infrastructure.

About DEL-AI

DEL-AI is Nurix’s discovery platform which employs advanced machine learning to enable all aspects of Nurix’s discovery engine, starting with DNA encoded library (DEL) hit-finding and degrader design, followed by automated chemistry synthesis and direct-to-biology screening and optimization, to rapidly generate degraders and degrader antibody conjugates (DACs) as new chemical entity drug candidates. By leveraging hundreds of billions of DEL compound binding signatures derived from thousands of DEL affinity screens collected from a diverse set of highly validated protein targets, Nurix’s DEL-AI platform can prospectively identify binders as starting points for drug discovery for virtually any pharmaceutically relevant target.

About Nurix Therapeutics, Inc.

Nurix Therapeutics is a clinical stage biopharmaceutical company focused on the discovery, development and commercialization of targeted protein degradation medicines, the next frontier in innovative drug design aimed at improving treatment options for patients with cancer and inflammatory diseases. Nurix’s wholly owned, clinical stage pipeline includes degraders of Bruton’s tyrosine kinase (BTK), a B-cell signaling protein, and inhibitors of Casitas B-lineage lymphoma proto-oncogene B (CBL-B), an E3 ligase that regulates activation of multiple immune cell types including T cells and NK cells. Nurix also is advancing multiple potentially first-in-class or best-in-class degraders and degrader antibody conjugates (DACs) in its preclinical pipeline. Nurix’s partnered drug discovery pipeline consists of preclinical stage degraders of IRAK4 and STAT6, as well as multiple additional programs under collaboration agreements with Gilead Sciences, Inc., Sanofi S.A. and Pfizer Inc., within which Nurix retains certain options for co-development, co-commercialization and profit sharing in the United States for multiple drug candidates. Powered by a fully AI-integrated discovery engine capable of tackling any protein class, and coupled with unparalleled ligase expertise, Nurix’s dedicated team has built a formidable advantage in translating the science of targeted protein degradation into clinical advancements. Nurix aims to establish degrader-based treatments at the forefront of patient care, writing medicine’s next chapter with a new script to outmatch disease. Nurix is headquartered in San Francisco, California. For additional information visit http://www.nurixtx.com.

Forward-Looking Statements

This press release contains forward-looking statements within the meaning of the U.S. Private Securities Litigation Reform Act of 1995 and other federal securities laws. Any statements contained herein that do not describe historical facts, including, but not limited to, statements regarding the potential advantages of Nurix’s DEL-AI platform and the potential benefits of Nurix’s DEL Foundation Model, including its potential to accelerate the discovery of novel drugs, are forward-looking statements that involve risks and uncertainties that could cause actual results to differ materially from those discussed in such forward-looking statements. Such risks and uncertainties include, among others, the risks described under the heading “Risk Factors” in Nurix’s Quarterly Report on Form 10-Q for the fiscal period ended February 28, 2025, and subsequent filings with the SEC. Any of these risks and uncertainties could materially and adversely affect Nurix’s business and results of operations, which could, in turn, have a significant and adverse impact on Nurix’s stock price. Nurix cautions you not to place undue reliance on any forward-looking statements, which speak only as of the date they are made. Nurix undertakes no obligation to update publicly any forward-looking statements to reflect new information, events or circumstances after the date they were made or to reflect the occurrence of unanticipated events.

Contacts:

Investors
Jason Kantor, Ph.D.
Nurix Therapeutics, Inc.
ir@nurixtx.com

Elizabeth Wolffe, Ph.D.
Wheelhouse Life Science Advisors
lwolffe@wheelhouselsa.com

Media
Aljanae Reynolds
Wheelhouse Life Science Advisors
areynolds@wheelhouselsa.com

FAQ

What is Nurix's DEL-AI platform and how does it speed up drug discovery?

Nurix's DEL-AI platform is a first-in-class foundation model trained on DNA encoded library data that can predict novel drug binders for disease targets. It accelerates drug discovery by enabling virtual screening of billions of compounds and identifying potential therapeutic candidates more efficiently than traditional methods.

How accurate is Nurix (NRIX) DEL Foundation Model in predicting drug compounds?

According to data presented at AACR 2025, Nurix's DEL Foundation Model accurately predicts experimental results and can identify novel binders with as little as 50% amino acid sequence similarity to training data. The model can also predict binders from chemical spaces not in the training set.

What makes Nurix (NRIX) DEL-AI platform unique for undruggable targets?

Nurix's DEL-AI platform is unique because it can predict novel binders for previously undruggable targets by combining a database of over 5 billion unique DEL compounds with machine learning models, enabling access to previously unreachable disease targets.

Who are the key partners in Nurix's (NRIX) DEL-AI development?

Nurix developed the DEL Foundation Model in collaboration with Loka, a Silicon Valley software firm, and Amazon Web Services (AWS), which provides enterprise-grade reliability and scalable infrastructure through AWS SageMaker and AWS managed MLflow.

What was revealed about Nurix (NRIX) at the AACR 2025 Annual Meeting in Chicago?

At AACR 2025 in Chicago (April 25-30), Nurix presented data showing their DEL Foundation Model can perform virtual DEL experiments to predict novel drug binders across the proteome, demonstrating its potential to transform drug discovery through AI-powered screening.