According to CoinMeta, recently, the research team led by Jiang Rui from the Department of Automation at Tsinghua University released a basic model for single-cell proteomics, ProtCell, and completed the paper "ProtCell: A foundation model for single-cell proteomics". This model utilizes approximately 380 million cells and 98 public datasets for pre-training and evaluation, enabling precise mapping of cell states and inference of unmeasured protein expressions, providing a crucial foundation for drug development and the construction of cutting-edge "virtual cells". The paper also reveals that the team successfully replicated the full process of pre-training ProtCell on Moore Threads' flagship integrated intelligent computing card MTT S5000. The main quantitative evaluation indicators showed only about a 0.2% difference compared to international mainstream GPU platforms, verifying the high capacity of Moore Threads' full-featured GPU (Universal GPU) and MUSA full-stack software ecosystem for advanced life science basic models.