Call for Pilot Project Proposals

The Center for AI-Driven Biomedical Research (AI4BIO) at Carnegie Mellon University’s School of Computer Science is pleased to announce its 2025 Call for Pilot Project Proposals. AI4BIO, established in fall 2024, aims to transform our understanding of living systems through AI/ML. Despite revolutionary advances in genomics, single-cell profiling, and high-throughput imaging, our ability to construct predictive, mechanistic models faithfully reflecting biological organization across scales—from molecular interactions to genome function, to cellular decision-making, to tissue architecture—remains profoundly limited. Bridging this gap requires a new generation of AI/ML methods that are not only data-driven, but also biologically grounded, interpretable, and experimentally actionable.

Part of AI4BIO’s mission is to foster collaborations to accelerate this transformation by supporting interdisciplinary projects that bring together computer scientists, engineers, and biomedical researchers across a wide range of disciplines to develop innovative approaches for modeling multiscale cellular systems and to tightly couple these models with next-generation experimental platforms, such as CMU’s automated science initiative. The goal is to establish a new paradigm for “closed-loop” science, in which AI models generate novel hypotheses that are rapidly tested and refined through automation, creating a virtuous cycle of discovery.

This pilot funding mechanism is designed to catalyze new collaborations, promote ambitious ideas, and seed larger-scale efforts that can further strengthen CMU’s leadership at the center of transformative advances in AI for biology and medicine.

Project Themes

We invite bold, interdisciplinary proposals that align with AI4BIO’s core research thrusts. Projects should seek to break new ground in how we model biological systems and/or how we conduct experiments—enabling a more predictive, mechanistic, and scalable understanding of life.

1. AI/ML for multiscale molecular and cellular modeling

Despite massive advances in genomic technologies and data generation, our understanding of how molecular components give rise to cellular behaviors—and how cellular interactions shape tissue function—remains fragmentary. We seek projects that develop AI/ML methods to bridge these gaps by learning from biological data at multiple spatial and temporal scales.

We are especially interested in proposals that leverage recent developments in generative AI, foundation models, and multimodal learning to uncover principles of genome regulation, cellular state transitions, and tissue-level organization. Projects should emphasize biological insight and testability, not just performance metrics. Successful projects may aim to:

• Develop models for predicting the impact of molecular variation on cellular function and phenotypes.
• Integrate multimodal data types (e.g., genomic and imaging) to understand structure and function of molecules, cells, tissues, and organs.
• Develop models that generalize across biological systems, including disease states.
• Explore emerging AI/ML architectures in the context of genomics and cell biology.

2. Integration of AI/ML with Automated Experimental Systems

A key goal of AI4BIO is to close the loop between computation and experimentation. We seek proposals that combine AI/ML model development with real-world experimental execution, particularly using automated laboratory infrastructure such as CMU’s Bakery Square Lab.

We particularly encourage high-risk, high-reward projects that showcase the power of model-experiment integration as a platform for accelerated scientific insight. These projects should embody the emerging paradigm of AI as a co-scientist—where models not only analyze data, but formulate, test, and refine biological hypotheses in coordination with automated experimentation. We encourage bold ideas that operationalize this concept, enabling AI to partner with human researchers in the full scientific loop.

We welcome proposals that:

• Use AI/ML to guide prioritization and optimization of experimental conditions for studying molecular and cell biology.
• Implement closed-loop systems between AI/ML models and robotic experimentation.
• Apply generative AI to design molecular constructs (e.g., DNA/RNA, proteins, cellular circuits) with targeted properties or behaviors.
• Explore frameworks that make AI models more actionable in experimental settings.

Eligibility, Collaboration Requirements, and Expectations

This pilot mechanism is designed to seed novel, high-risk, and interdisciplinary research directions that would not be pursued under existing funding. Projects that primarily extend ongoing collaborations or already funded research are unlikely to be considered competitive. Proposals must clearly articulate how the work represents a substantial departure from current efforts and opens new intellectual territory.

To ensure broad engagement and cross-disciplinary innovation, the following eligibility and collaboration requirements apply:

• Each project must be led by at least two faculty members from distinct research groups, preferably from different departments or colleges.
• At least one faculty lead must have a primary appointment in the School of Computer Science or be based in another CMU unit but demonstrably advancing the state of computational methods in their discipline. The intent is to ensure deep computational innovation, while fostering cross-disciplinary collaborations.
• Proposals should demonstrate a true collaborative effort, with meaningful contributions from all participating investigators.

Awardees will also be expected to actively contribute to the AI4BIO community. Awardees must:

• Participate in center-sponsored seminars, symposia, and working groups to share progress and stimulate new collaborations.
• Present their work at least once during the AI4BIO seminar series or internal workshop.
• Acknowledge AI4BIO pilot support in any resulting publications, talks, or proposals.
• Engage in future multi-investigator grant proposals organized by the AI4BIO Center.
• Participate in reviewing processes for potential future solicitations.
• Submit a short final report at the end of the funding period outlining scientific progress, outcomes, and future directions.

 Funding Scope

Each selected project will receive up to $100,000 in total funding for a one-year period. The award is intended to provide targeted support for launching new, high-impact research directions.

Funding may be used for:

• One Ph.D. student FTE for one year of tuition and stipend (required). This may be distributed across two students at 50% effort each, if better aligned with the proposed research activities.
• Research-related expenses, such as laboratory consumables, equipment, data acquisition, cloud computing, or conference travel directly related to the project.

Important constraints:

• No faculty summer salary or PI effort compensation will be provided under this award.
• No cost extensions are not allowed.

Start Date and No Extension Policy

Awards will begin Spring 2026, and must be completed within 12 months. No-cost extensions or delays in project start will not be permitted. This is intended to ensure rapid progress, public visibility, and support for future center-wide funding efforts.