CASP, DREAM, OpenADMET, Sage Bionetworks, and CACHE/Conscience join forces to foster synergies among efforts to benchmark and accelerate AI-driven science

TORONTO: Today marks the official launch of BEACON (Benchmarking, Evaluation, and Assessment Consortium for Science), a new consortium dedicated to establishing a comprehensive and adaptable framework for validating research and predictions on complex or emerging biological systems. The consortium brings together leaders across disciplines to conduct open evaluations of predictive models in biology and medicine, coordinate community efforts to advance reproducible science, and guide AI-driven innovation in disease biology and drug discovery, ultimately accelerating new directions in research and catalyzing the development of safer, more effective treatments with real-world impact.

Conscience, a non-profit focused on enabling drug discovery and development through open collaboration, is acting as the central coordinator for the initiative, facilitating collaboration and ensuring the consortium meets its goals of transparency, rigour, and broad scientific and community engagement.

BEACON unites five organizations with deep expertise and proven track records in benchmarking and critical assessment: CASP (Critical Assessment of Structure Prediction), DREAM (Dialogue on Reverse Engineering Assessment and Methods), Sage Bionetworks, OpenADMET and CACHE/Conscience. The consortium aims to increase the collective impact of critical assessment approaches across the scientific community, strengthen rigour and reproducibility in research, and create a comprehensive, adaptable benchmarking framework that advances AI and computational approaches to disease research and drug discovery. It also focuses on developing innovative solutions for benchmarking foundation models and AI agents. The consortium welcomes participation from other researchers, institutions, and organizations committed to advancing open and reproducible science.

BEACON’s initiatives will include a benchmarking think-tank to promote methodological coherence and thought leadership, and an open platform that runs challenges, engages communities, and provides easy access to data and methods for accelerated collective learnings, among other tasks.

Founding Governance Committee members of BEACON include:

• John Moult - Co-Founder & Chair, CASP; Co-Founder & Initial Co-Chair, CAGI (Critical Assessment of Genome Interpretation); and Professor, Department of Cell Biology & Molecular Genetics, University of Maryland
• Julio Saez-Rodriguez - Head of Research, EMBL-EBI; Director, DREAM Challenges; Professor, Computational Biomedicine
• Pablo Meyer – Senior Research & Scientist Manager for IBM Research, Director and Vice-Chair of DREAM Challenges
• Gustavo Stolovitzky - Founder, Chair, DREAM Challenges; Director, Biomed Data Science Hub, NYU Langone Health; Professor, Pathology Department, NYU School of Medicine
• Luca Foschini - President, Sage Bionetworks
• Aled Edwards - Co-Founder & Chief Scientific Officer, Conscience; CEO, Structural Genomics Consortium; Temerty Nexus Chair of Health Technology and Innovation, University of Toronto and University Health Network
• Patrick Walters - Chief Scientist, OpenADMET
• Peng Fu - CEO Conscience

Members of the consortium will be participating in the MAINFRAME Symposium on AI-Driven Small-Molecule Drug Discovery in Barcelona from March 18-19, 2026. The Symposium aims to bring together the global community working at the intersection of artificial intelligence, computational chemistry, and small-molecule drug discovery. Conscience, together with several founding partners of BEACON, will lead a dedicated session on benchmarking to formally introduce the consortium, outline its approach to rigorous and open evaluation of AI methods, and engage directly with the machine learning research community. The session will celebrate BEACON’s public launch and provide an opportunity for researchers to offer feedback and help shape the initiative’s early priorities.

Quotes

“Biology is generating extraordinary amounts of data, but our ability to interpret and critically assess that information has not kept pace, contributing to replication failures and research misdirection,” says John Moult, Co-founder and Chair of CASP. “CASP and other initiatives like it have shown that open, community-driven science, combined with rigorous and independent assessment, can produce validated answers to complex biological problems at scale. BEACON applies those same principles, integrating open collaboration, advanced AI, and critical evaluation to identify gaps, clarify what is reliable, and inform more effective directions for future research.”

“DREAM is extremely pleased to join BEACON,” says Gustavo Stolovitzky, Founder and Chair of the DREAM Challenges. “The field of critical assessment has become increasingly fragmented, and we hope to attract other kindred organizations that promote crowdsourced, collaborative competitions to join BEACON in advancing the principles of open, collaborative, and reproducible science.”

"Rigorous benchmarking helps define the best approaches to analyze data, including the data publicly available at EMBL-EBI. BEACON will thus accelerate research across biomedical domains” says Julio Saez-Rodriguez, Head of Research, EMBL-EBI and Director of the DREAM Challenges.

“AI is rapidly advancing in biology, increasingly acting as a co-pilot for hypothesis generation and analysis. However, rigorous, open evaluation is necessary to ensure these tools are reliable,” states Luca Foschini, President of Sage Bionetworks. “Sage has long been a platform for open benchmarking and challenge-driven science across biomedical fields. BEACON builds on this proven model, offering a coordinated way to measure progress and establish trust in AI-enabled biology.”

"To advance machine learning in drug discovery, we must move beyond an obsession with algorithms and address the data gap,” says Patrick Walters, Chief Scientist at OpenADMET. “Data is the most critical component of any model, yet it is often the most overlooked. BEACON represents a concerted effort to deliver the high-quality, consistent data required to turn machine learning potential into a reality for drug discovery.”

“AI can only advance as fast as our ability to measure it. Benchmarking and critical assessment are the cornerstone of modern science and necessary for objective validating progress,” says Peng Fu, CEO of Conscience. “By coordinating community efforts with shared core values of open science, teamwork, and impact, we can establish comprehensive, adaptable frameworks that drive the development of AI and foundation models and deliver meaningful impact across the sciences.”

Conscience, together with BEACON’s founding partners, is calling on the biomedical community, including academia, industry, funders, and policymakers, to support benchmarking as the foundation for the responsible use of AI in drug discovery. BEACON will provide the infrastructure and expertise, but the consortium is calling for industry commitment and collaboration to sustain its efforts. To learn more, please visit conscience.ca/beacon.

About Conscience

Conscience is a non-profit focused on enabling drug discovery and development in areas where open sharing and collaboration are key to advancement towards accessible treatments. It does so by encouraging and funding the open sharing of knowledge and tools, the use and improvement of artificial intelligence, and the development of policies that break down barriers of traditional drug development. Powered by a network that includes academics, industry, technologists, policy experts, and public support, Conscience seeks to drive innovation by turning drug discovery and development into a team sport. Its open science model brings unique value in areas where market solutions are limited, offering alternatives to traditional intellectual property models to make new accessible medicines so no one is left behind. Through key initiatives, such as its DMOS (Developing Medicines through Open Science) program and CACHE (Critical Assessment of Computation Hit-finding Experiments) Challenges, Conscience is accelerating the path to treatments for those who need them most. For more information, visit conscience.ca.

Fonte: Business Wire