Methods and systems for tissue informed differentially methylated region analysis
A method using a universal panel of genomic regions and advanced sequencing techniques for ctDNA analysis addresses the challenge of low ctDNA fraction in blood, enabling accurate cancer detection and monitoring with high sensitivity and specificity.
Patent Information
- Application Number
- PCT/US2025/016672
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-10-24
- Filing Date
- 2025-02-20
- Publication Date
- 2025-08-28
AI Technical Summary
Existing methods for analyzing circulating tumor DNA (ctDNA) as a non-invasive biomarker for cancer detection face challenges due to the low fraction of ctDNA in blood-derived samples and the need for improved methods to accurately identify differentially methylated regions (DMRs) for early cancer detection and monitoring.
A method involving the use of a universal panel of genomic regions to enrich and analyze nucleic acid molecules from samples, including sequencing at a depth of up to 50 million single reads, and comparing these regions to reference sets to generate methylation scores for cancer detection and monitoring, utilizing techniques such as Support Vector Machine and logistic regression.
This approach enables accurate detection and monitoring of cancer with high specificity, allowing for the identification of minimal residual disease and guiding therapy decisions, with sensitivity and specificity up to 90%.