Human Disease Blood Atlas Maps Protein Fingerprints Across 59 Conditions

An international team of more than 100 co-authors has created a comprehensive Human Disease Blood Atlas, mapping how thousands of proteins in our blood shift across 59 diseases and healthy life stages. Published today in Science, “A human pan-disease blood atlas of the circulating proteome,” profiled more than 8,000 individuals for the blood levels for up to 5,400 proteins, revealing that each disease leaves a distinct molecular “fingerprint” in blood, and that many conditions also share common signatures of inflammation and tissue injury. The full, interactive dataset is now live on the open-access Human Protein Atlas.

“By comparing 59 diseases side by side, we can separate universal ‘alarm bells’ of inflammation from truly disease specific signals, which is a crucial step for building blood tests that work in the clinic,” said Professor Mathias Uhlén, senior author and director of the Human Protein Atlas. “Every day, around 70 new biomarker studies are published worldwide, but most compare disease against controls,” says María Bueno Álvez, researcher at SciLifeLab and first author on the paper. “Because many proteins show variability across multiple conditions, such narrow comparisons often produce results that cannot be reproduced, contributing to the wider reproducibility crisis in today’s science.”

“With this study, we thus demonstrate how a pan-disease approach can help overcome that challenge,” adds Professor Mathias Uhlén, senior author of the study and director of the Human Protein Atlas.

Key findings

• Distinct disease fingerprints and shared biology: Each of the 59 diseases showed characteristic protein profiles. Many proteins that rise in cancer or autoimmunity also rise in infections, reflecting shared inflammatory pathways, while other patterns clustered by organ systems (for example, liver related conditions). This dual view helps focus on truly disease specific markers.
• AI-based analysis separates universal from disease specific signals: Machine learning models showed accurate disease classification and identified disease-specific signatures in the plasma proteome, information critical for building panels that won’t misclassify patients in real world settings.
• Your baseline matters: Longitudinal profiling in healthy adults showed that each person has a stable, individual blood protein baseline over years, a molecular fingerprint of wellness that future healthcare could compare against to flag early deviations. “The best control for your health is you,” said Álvez.
• Puberty reshapes the blood proteome: Tracking 100 individuals at ages 4, 8, 16, and 24, the team observed dramatic, sex specific shifts during puberty before levels stabilize in adulthood, creating one of the most detailed maps to date of how development rewires the circulating proteome.
• Prediction of biological age: Models predicting the chronological age of the individual based simply on the blood proteome profile showed good performance. Individuals predicted to have higher or lower biological age compared to the chronological age stood out as outliers, an observation the authors note is interesting for further study.

Signals relevant to early cancer detection

Reference: Álvez MB, Bergström S, Kenrick J, et al. A human pan-disease blood atlas of the circulating proteome. Science. 0(0):eadx2678. doi: 10.1126/science.adx2678

This article has been republished from the following materials. Note: material may have been edited for length and content. For further information, please contact the cited source. Our press release publishing policy can be accessed here.