MRD Detection Using Trinucleotide Error Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting minimum residual disease (MRD) in cancer patients are inaccurate due to varying sequencing errors that are not adequately accounted for, leading to false positives and negatives in the identification of cancer-associated mutations.
Innovation Solution
A computer-implemented method that determines sequencing error rates within a biological sample by monitoring trinucleotide context (TNC) error rates, grouping these rates, and using a statistical hypothesis test to differentiate between actual mutations and sequencing errors, thereby improving the accuracy of MRD detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional sequencing methods are used to detect mutations, then the detection process is simple, but the accuracy is low due to sequencing errors causing false positives and negatives
Solution Approach 1:
The patent segments the mutation detection process into distinct phases: (1) determining TNC error rates from sequencing data, (2) grouping TNC error rates into categories, (3) determining TNC group error rates, and (4) comparing actual mutations against expected error rates. This segmentation allows systematic handling of sequencing errors while maintaining detection accuracy.
Solution Approach 2:
The patent performs preliminary error rate characterization before final mutation detection. By first determining TNC error rates and grouping them, the method establishes a baseline of expected sequencing errors that is then used to evaluate actual mutations, enabling more accurate distinction between true mutations and sequencing artifacts.
2Reliability
If sequencing error rates are not adequately accounted for, then the detection method is straightforward, but false positives and negatives increase
Solution Approach 1:
The patent implements feedback by using determined TNC error rates to adjust and refine mutation detection thresholds. The error rate information feeds back into the detection process, allowing dynamic adjustment of significance thresholds based on actual sequencing performance, thereby improving reliability while managing complexity through iterative refinement.
Solution Approach 2:
The patent changes parameters by transitioning from fixed significance thresholds to dynamic thresholds based on determined TNC error rates. By adjusting detection parameters according to measured error rates, the method adapts to varying sequencing conditions and improves reliability without requiring overly complex additional hardware.
3Measurement precision
If TNC error rates are determined and grouped for each sequence read, then sequencing error accuracy improves, but computational time and resources increase
Solution Approach 1:
The patent merges individual TNC error rates into grouped categories (TNC group error rates), combining multiple individual measurements into consolidated rate estimates. This merging reduces the computational burden of evaluating each TNC separately while maintaining precision through aggregated statistical analysis, thereby improving throughput without sacrificing accuracy.
Solution Approach 2:
The patent applies partial action by focusing error rate determination on relevant TNC contexts rather than analyzing every possible nucleotide position in depth. By concentrating computational resources on trinucleotide contexts that are most informative for error characterization, the method achieves sufficient precision without excessive computational overhead that would reduce throughput.
Data Source
AI summary
The present disclosure describes techniques for determining an indication of minimum residual disease (MRD) in a subject. The indication of MRD may be determined based on sequencing data from a biological sample of the subject. These techniques are performed in part by determining sequencing error and an indication MRD from the same biological sample using the same set of sequencing data.