The international team profiled the blood concentrations of up to 5,416 proteins in 8,262 individuals, including healthy donors and patients with 59 different diseases, creating a pan-organism atlas of the circulating proteome[1]. The study identified proteins associated with age, gender and body mass index, while also determining disease-specific protein signatures[1]. Several protein changes appeared to be common across different diseases (particularly in infections and liver disease), while some disease groups showed clearly distinguishable patterns by organ system[3][5]. The authors used machine learning to separate universal inflammatory signals from truly specific biomarkers, which improved disease classification within the atlas[3][5]. Long-term (longitudinal) samples have shown that the individual adult proteome tends to be stable and that concentrations of many proteins reach "adult" levels by about 16 years of age[6]. Some protein profiles were already changing prior to the clinical diagnosis of cancer, suggesting the potential of proteomics in early cancer detection[5]. The data is being made available as an open online resource within the Human Protein Atlas for further research and biomarker validation[1][6].