CytoReason has unveiled LINA, an AI agent designed to support pharmaceutical research and development, at the 2026 JP Morgan Healthcare Conference. LINA acts as a computational biology assistant, interpreting molecular, clinical, and patient-level data using CytoReason’s AI disease models combined with NVIDIA NIM microservices.

The AI agent is built on CytoReason’s framework and trained on hundreds of pharma R&D questions, allowing it to deliver grounded analysis plans, generate validated and reproducible code, and provide clear biological explanations across genes, pathways, cell types, and patient subgroups. By anchoring responses to CytoReason’s mechanistic disease models, LINA avoids the hallucinations commonly seen with general-purpose language models.

Running on NVIDIA accelerated computing and the Qwen3-Next-80B-A3B-Instruct model hosted as NVIDIA NIM microservices, LINA provides optimized, containerized infrastructure for secure deployment, fast inference, and integration into existing pharma R&D workflows. This enables research teams to perform complex biological analyses at scale, significantly reducing analysis cycles.

CytoReason is collaborating with leading pharmaceutical companies to validate LINA’s capabilities. In partnership with Takeda, the company published data at the United European Gastroenterology (UEG) Congress showing that TYK2-dependent gene signature enrichment is associated with both ulcerative colitis and Crohn’s disease. These findings highlight LINA’s ability to generate new insights for target discovery and support evidence-based decision-making in drug development.

Prof. Shai Shen-Orr, co-founder and chief scientist of CytoReason, said: “At CytoReason, we set out to build the computational backbone of biological intelligence. LINA is the realization of that vision. With computational disease models, molecular and clinical data, and advanced AI infrastructure working together, LINA empowers scientists to engage directly with disease biology and get grounded, biological answers. We’re proud to work with NVIDIA and partners like Takeda to bring this technology to life and to reshape how R&D teams make decisions.”

The platform also generates visualizations, summaries, and end-to-end reports, enabling faster, more confident decisions at critical points throughout the drug development lifecycle. By supporting scalable, multi-agent scientific workflows, LINA aims to improve the speed, accuracy, and transparency of R&D decisions, positioning computational disease models as a core infrastructure for precision medicine.