Current liquid biopsies struggle to pinpoint where disease signals originate, limiting their use. This new “cf-EpiTracing” platform overcomes that by capturing detailed epigenetic fingerprints from trace blood samples. It can identify the specific tissues driving a disease, distinguish lymphoma subtypes, and predict patient outcomes better than existing clinical tests, paving the way for earlier, more precise non-invasive diagnoses.

In the field of early diagnosis and screening for colorectal cancer, cf-EpiTracing has delivered impressive results. By integrating multimodal epigenomic features from cell-free chromatin and leveraging machine learning algorithms, cf-EpiTracing reaches an accuracy rate of up to 97.6% in training group samples, and remains robust at 92.2% in independent validation group samples.