Research on thousands of proteins in a drop of blood shows that proteins can predict the occurrence of various diseases. This study is part of an international research project jointly conducted by GlaxoSmithKline, Queen Mary University of London, University College London, the University of Cambridge, and the Berlin Institute for Health Research at the Sharrit Medical School in Germany. It was published on the 22nd in the journal Nature Medicine. This project is the largest proteomic study to date. The team used data from the UK Biobank Pharmaceutical Proteomics Project to measure approximately 3000 plasma proteins collected from over 40000 randomly selected participants in the UK Biobank. Protein data is associated with participants' electronic health records. Using advanced analytical techniques, the team accurately identified 5 to 20 "features" of the most important proteins for predicting each disease. The research report states that protein features can predict the onset of 67 diseases, including multiple myeloma, non Hodgkin lymphoma, motor neuron disease, pulmonary fibrosis, and dilated cardiomyopathy. The performance of protein prediction models is superior to models based on standard clinical record information. In most cases, the prediction performance based on blood cell count, cholesterol, renal function and diabetes test (glycosylated hemoglobin) is not as good as the protein prediction model. The research team stated that the performance of several protein features is similar or even better than proteins that have been tested, as they have screening potential, such as prostate-specific antigen for prostate cancer. Therefore, the newly discovered protein features may provide early detection and ultimately improve prognosis for many diseases, including severe diseases such as multiple myeloma and idiopathic pulmonary fibrosis. A key challenge in drug development is to identify the patients most likely to benefit from the new drug. This study suggests that using large-scale proteomics techniques to identify high-risk individuals for multiple diseases has promising prospects. Revealing future disease risks, such as the likelihood of heart attack and stroke, can save thousands of lives. But how to measure and ensure accuracy has always been a challenge. This study has brought new possibilities for predicting various diseases. In addition to common high-risk diseases, there are also rare diseases that may take months or even years to be diagnosed. It should be noted that it is the larger databases and more precise analysis methods, combined with corresponding technologies, that continuously deepen people's understanding of biology and diseases. (New Society)