Researchers at Peking University have developed cf-EpiTracing, an advanced platform that detects and traces diseases using just 50 microliters of human plasma—about one drop of blood. The study appeared in Nature on March 4, 2026, under the leadership of Professor He Aibin from the College of Future Technology and Professor Jing Hongmei from the Department of Hematology at PKU Third Hospital.
Why It Matters
Traditional liquid biopsies often fail to identify the origin of disease signals in blood samples, restricting their clinical applications. cf-EpiTracing addresses this limitation by capturing detailed epigenetic profiles from minimal blood volumes. The platform pinpoints specific tissues fueling diseases, differentiates lymphoma subtypes, and outperforms standard tests in predicting patient outcomes, enabling earlier and more accurate non-invasive diagnostics.
Key Findings
For early colorectal cancer detection and screening, cf-EpiTracing combines multimodal epigenomic data from cell-free chromatin with machine learning, achieving 97.6% accuracy in training samples and 92.2% in independent validation sets.
In diffuse large B-cell lymphoma cases, the platform reveals elevated CD34-positive cell signals in plasma, indicating potential bone marrow involvement and disease severity. These insights support improved subtyping and tailored treatment approaches.
Future Implications
Upcoming advancements will merge cf-EpiTracing with additional cell-free analyses, including DNA methylation, mutations, and chromatin structure. This multi-omics strategy could deliver superior precision for diagnosing complex conditions, tracking disease progression, and evaluating treatments in large-scale patient studies, reshaping non-invasive diagnostics across clinical fields.
