Tumor-Informed MRD Assay Error Correction via Proximal Controls
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting circulating tumor DNA (ctDNA) in cancer patients face challenges due to low levels of ctDNA and high background error rates, which hinder accurate detection and estimation of tumor fraction.
Innovation Solution
The proposed method involves post-sequencing error-correction and the use of internal controls proximate to variant sites to improve the sensitivity and accuracy of tumor-informed minimal residual disease (MRD) assays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard sequencing methods are used to detect ctDNA, then the detection process is simple and quick, but the detection sensitivity is low due to high background error rates
Solution Approach 1:
The patent applies preliminary action by performing error correction computations before final ctDNA detection. The system pre-calculates error rates for each somatic variant using control sites, and uses these pre-computed error rates to set detection thresholds. This preliminary error characterization enables the system to distinguish true ctDNA signals from background errors more effectively, improving detection sensitivity without requiring complex hardware modifications.
Solution Approach 2:
The patent introduces control sites as intermediary elements that do not directly detect ctDNA but serve as mediators to characterize background error rates. These control sites are sequenced alongside target somatic variants, and their error patterns are used to model and correct errors in the actual detection process. This intermediary approach allows the system to account for context-specific errors without adding direct detection complexity to the target sites.
2Measurement precision
If tumor fraction estimation is performed using current methods, then the process is straightforward, but the estimation accuracy is poor due to inability to account for context-specific errors
Solution Approach 1:
The system performs preliminary error rate calculations for each somatic variant using control sites before tumor fraction estimation. These pre-computed error rates are stored and reused for multiple samples, avoiding redundant computations. This preliminary action separates the error characterization step from the estimation step, enabling accurate context-specific error correction while reducing per-sample computation time.
Solution Approach 2:
The patent implements feedback by using control site data to continuously refine error rate estimates for each somatic variant. The error rates calculated from control sites are fed back into the tumor fraction estimation algorithm, allowing the system to adjust its detection thresholds and confidence calculations based on empirically observed error patterns. This feedback mechanism improves estimation accuracy by accounting for context-specific biases in different genomic regions.
3Reliability
If minimal residual disease assays are performed without error correction, then the assay is fast and simple, but the reliability of positive detections is low due to false positives from background errors
Solution Approach 1:
The patent uses control sites as intermediary elements to mediate between the raw sequencing data and the final detection decision. These control sites are sequenced in the same manner as target sites but are known to be error-free (or have known error rates). By comparing target site signals against control site-derived error models, the system can reliably distinguish true positive detections from false positives without requiring complex computational corrections at each detection step.
Solution Approach 2:
The patent replaces mechanical/error-prone physical processes with computational error correction. Instead of relying on physically perfect sequencing reactions, the system uses computational models that calculate expected error rates based on control site data. This substitution allows the system to achieve high reliability by using information theory and statistics to correct for inherent limitations in the physical sequencing process, rather than attempting to eliminate errors at the physical level.
Data Source
AI summary
Described herein are methods for improving sensitivity and estimation of tumor-informed MRD assays and reducing error rates by performing post-sequencing error-correction and error correction using internal controls that are proximate to the variant sites.