Researchers have developed sparse prediction models combining plasma proteins and clinical data to predict the 10-year risk of 218 diseases. The study, involving over 41,000 participants, highlights the enhanced predictive power of proteomic signatures.