Methods and systems for high-depth sequencing of methylated nucleic acid
Inventors
Ariazi, Eric • Weinberg, David • Hogan, Greg • St. John, John • Pearson, Michael
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
Methods and systems provided herein address current limitations of bisulfite-based methylation sequencing by improving the quality and accuracy of nucleic acid methylation sequencing and uses thereof for detection of disease. Methods that include minimally-destructive conversion methods for methylation sequencing as well as specialized UMI adapters provide for improved quality of sequencing libraries and sequencing information. Greater accuracy and more complete methylation-state information permits higher quality feature generation for use in machine learning models and classifier generation.
Core Innovation
The disclosed invention concerns cancer detection from subject-derived nucleic acid, including cell-free DNA, by combining conversion-tolerant adapter design and enzymatic cytosine conversion with targeted analysis of CpG, CHG, or CHH loci. Sequence information is compared with a reference and evaluated by a trained machine-learning model to detect cancer. The supplied summary also describes unique molecular identifiers and duplex error correction as supporting sequencing features.
The supplied partial content does not include the background section, so the specific problem statement from that section is not explicitly described. It states that the approach improves DNA preservation, coverage, uniformity, and accuracy relative to bisulfite methods. It also reports higher sequencing-library yield for a conversion-tolerant adapter than for a 5-methylcytosine-containing adapter under conversion conditions.
Claims Coverage
The provided relevant claims identify one independent claim directed to a method of detecting cancer in a subject, reciting five inventive features.
Conversion-tolerant adapter
The claim recites ligating a conversion-tolerant nucleic acid adapter to a subject-derived nucleic acid molecule.
Enzymatic cytosine conversion
The claim recites enzymatically converting unmethylated cytosines to uracils and amplifying the converted nucleic acid molecules.
Targeted enrichment of selected loci
The claim recites using probes to enrich sequences corresponding to a pre-identified panel of CpG, CHG, or CHH loci.
Sequencing and reference comparison
The claim recites sequencing enriched nucleic acid molecules and comparing their sequence to a reference sequence of the pre-identified panel.
Machine-learning cancer detection
The claim recites processing sequence information with a trained machine-learning model configured to detect cancer. The training data comprise samples from subjects with advanced adenoma and samples from subjects without advanced adenoma.
The independent claim combines conversion-tolerant adapter use and enzymatic cytosine conversion with targeted locus analysis, sequencing, reference comparison, and trained machine-learning cancer detection.
Stated Advantages
The disclosure states improved DNA preservation, coverage, uniformity, and accuracy compared with bisulfite methods.
The disclosure reports higher sequencing-library yield for a conversion-tolerant adapter than for a 5-methylcytosine-containing adapter under conversion conditions.
Documented Applications
Cancer detection and monitoring, including colorectal cancer and minimal residual disease.
Interested in licensing this patent?