SHANGHAI, CHINA – MobiDrop and INFevo have announced a joint strategic R&D partnership aimed at advancing the emerging field of virtual cells. The collaboration brings together MobiDrop's large-scale life science data production capabilities and INFevo's foundational virtual cell modeling expertise to build an integrated, closed-loop "Data–Model–Experiment" R&D infrastructure—transforming virtual cells from a scientific concept into a sustainable, scalable, and industrially applicable platform.

A New Paradigm for AI-Driven Life Sciences

As an AI-era life science infrastructure provider, MobiDrop harnesses proprietary single-cell technologies, automated laboratory platforms, and intelligent analytical systems to generate high-quality, multi-omics, AI-ready biological data at scale. This continuous data production forms a stable and reliable foundation for model training and ongoing algorithmic refinement.

INFevo, a pioneer in digital biology and computational life sciences, centers its efforts on a proprietary virtual cell foundational model. The company is developing an intelligent research system tailored for drug discovery, embedding model prediction, experimental design, validation, and result feedback into a unified, end-to-end workflow. This iterative loop enables the model to evolve with each cycle of learning and experimentation.

Through this partnership, the two companies will bridge the critical gap between real-world biological experimentation and AI modeling. Experimental data will continuously inform and refine model learning, while model predictions will actively guide and optimize experimental design—forming a virtuous cycle in which "experiments generate data, models propose hypotheses, and experiments validate models." The result is a more robust foundation for the industrialization of virtual cell technologies.

Technology Co-Innovation: Building the "Data–Model" Closed Loop for Virtual Cells

The joint R&D initiative will focus on multi-omics data integration and AI model development for virtual cells, with immediate applications in innovative drug discovery and precision medicine. Under the collaboration framework:

· MobiDrop will design and generate experimental datasets tailored to model training requirements.

· INFevo will leverage model predictions to propose and refine experimental protocols in return.

Operating under a unified data specification, both parties will engage in bidirectional iteration and cross-validation. In this new paradigm, experiments are no longer merely a final validation step—they become an integral source of model training data. Likewise, models transcend their traditional role as predictive tools and become active participants in shaping the next generation of experiments.

The partnership will also explore joint initiatives in data governance, model benchmarking, quality standards, and application guidelines, with the goal of establishing industry-wide data standards, evaluation frameworks, and best practices—delivering replicable and scalable foundational capabilities for the broader virtual cell ecosystem.

Ecosystem Co-Building: Fostering an Open and Collaborative Virtual Cell Innovation Community

Beyond joint R&D, MobiDrop and INFevo are committed to nurturing a thriving innovation ecosystem around virtual cells. Through collaborative research programs, open datasets, technical white papers, and industry forums, the two companies will actively promote the open and synergistic development of data, models, and experimental capabilities.

The partners will engage with academic research institutes, pharmaceutical companies, healthcare institutions, and other industry stakeholders to collectively accelerate the advancement of virtual cell technologies. Looking forward, the two companies plan to progressively open their "Data–Model–Experiment" closed-loop capabilities to the wider industry, accelerating the real-world deployment of virtual cells in drug discovery, precision medicine, biomanufacturing, and beyond—while providing continuously evolving infrastructure support for AI-powered life science innovation.

A Strategic Milestone Toward the Future of Life Sciences

This strategic collaboration marks a pivotal step in MobiDrop's mission to build next-generation life science infrastructure for the AI era. It also represents a meaningful practice in the deep convergence of high-quality biological data and virtual cell modeling.

The life sciences are undergoing a fundamental transformation—moving away from traditional, experience-driven R&D models and toward a new paradigm defined by data-driven discovery, model-based prediction, and experimentally validated iteration. Moving forward, MobiDrop will continue to partner with ecosystem collaborators to strengthen the "Data–Model–Experiment" infrastructure, accelerate the industrial adoption of virtual cell technologies, and deliver a lasting data foundation and innovation engine for drug discovery, precision medicine, and biomanufacturing.