Biological activities are inherently highly complex. Even single‑cell model organisms such as Escherichia coli and Saccharomyces cerevisiae exhibit marked systemic complexity in their biological processes. Human diseases represented by cancer and Alzheimer’s disease possess even higher‑dimensional spatiotemporal complexity. Reliance on a single disease target or pathway can hardly achieve full mechanistic dissection, precise diagnosis and effective intervention — this underlies why many complex human diseases still lack effective diagnosis and treatment. Proteins are the primary effector molecules of life activities and core targets for drug action. Thousands of proteins assemble into sophisticated molecular machineries that carry out physiological functions across multiple spatiotemporal scales at the cellular, tissue and organismal levels, forming highly intricate proteomic systems. Nevertheless, our current understanding of the spatiotemporal mechanisms governing proteome function remains limited. To address these challenges, we are committed to establishing a spatiotemporal proteomics methodology system, including foundational models for protein identification and quantification, spatial proteomics, and perturbation proteomics. These deliver core technical support for developing AI‑driven virtual cell models. Leveraging model systems including Escherichia coli and Saccharomyces cerevisiae, we integrate multi‑modal and multi‑omics datasets and perform systematic analysis with large‑language models. We explore systematic approaches for constructing virtual cell models to gain a holistic understanding of complex biological principles, thereby advancing the diagnosis and treatment of complex diseases such as cancer and Alzheimer’s disease, alongside the systematic engineering of model organisms.
