cfDNA Fragment Marker Modeling for Early Multi-Cancer Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cancer screening methods are invasive, painful, costly, and have low compliance, limited detectability for early-stage cancer, and high false positive/negative rates, lacking effective standard screening methods for various cancers.
Innovation Solution
A multi-cancer early detection model using low-coverage whole-genome sequencing of cell-free DNAs, analyzing six differential features, and integrating convolutional neural networks with multiple algorithms to construct a model for non-invasive, accurate detection and tissue-of-origin tracing of multiple cancers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional screening methods such as MRI, CT, endoscopy, or tissue biopsy are used, then cancer can be detected, but the methods involve high radioactivity, invasive procedures, pain, difficulty in sampling, and low compliance
Solution Approach 1:
The patent replaces mechanical/invasive screening methods (endoscopy, biopsy, imaging) with a molecular biology-based liquid biopsy approach. Cell-free DNA fragments are extracted from blood plasma and analyzed through next-generation sequencing, substituting physical intrusion with biochemical analysis that is non-invasive and highly compliant
Solution Approach 2:
The patent uses cell-free DNA fragments as an intermediary substance that carries cancer-related genetic information from tumor cells into the bloodstream. These cfDNA fragments serve as mediators that allow indirect detection of cancer without directly contacting or invading the tumor tissue, enabling non-invasive screening while maintaining high detection accuracy
2Measurement precision
If tissue biopsy is performed to obtain cancer samples, then diagnostic information can be obtained, but the method involves difficult sampling, incomplete sampling due to tumor heterogeneity, and high false positive/negative rates
Solution Approach 1:
The patent makes the screening method universally applicable to multiple cancer types simultaneously. By analyzing cell-free DNA fragments circulating in the bloodstream, a single blood draw can detect genetic alterations associated with various cancers (lung, liver, colorectal, pancreatic, etc.), eliminating the need for multiple specialized procedures or complex sampling strategies for different tumor locations
Solution Approach 2:
The patent extracts and analyzes specific cancer-related genetic information from the complex mixture of cell-free DNA in blood plasma. Through next-generation sequencing and bioinformatics analysis, the method extracts tumor-derived genetic markers (mutations, copy number variations, methylation patterns) from the background of normal DNA, achieving high diagnostic precision without the sampling difficulties of tissue biopsy
3Reliability
If standard screening methods are used for breast cancer, lung cancer, and prostate cancer, then detection can be performed, but compliance remains low and detectability for early-stage cancer is limited
Solution Approach 1:
The patent performs preliminary detection of cancer-related genetic alterations in cell-free DNA before clinical symptoms appear or tumors become detectable by traditional imaging. By analyzing molecular markers in blood plasma, the method can identify early-stage cancer signals that precede anatomical changes, enabling earlier intervention while maintaining high compliance through non-invasive blood sampling
Data Source
AI summary
The present disclosure relates to an application of gene markers in multi-cancer early detection, a method for constructing an early detection model, and a detection device. In the present disclosure, low-coverage whole-genome sequencing is conducted on cell-free DNAs (cfDNAs) from a plasma sample, and according to high-throughput sequencing results, six differential features of the cfDNA fragments are analyzed for each cancer. Then the training and modeling are conducted with a convolutional neural network to allow the early detection of a plurality of cancers at a low sequencing depth. Then the training and modeling are conducted with a generalized linear model (GLM), a gradient boosting machine, a random forest model, a deep learning model, and an extreme gradient boosting model, and staking is conducted with a GLM to construct a multi-feature algorithm, to allow the tissue-of-origin-based detection of cancers.