Genomic Alteration Detection via Replicate Statistical Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current genetic mutation detection methods face challenges in sensitivity and specificity, particularly at low allelic frequencies (AF), often resulting in high false positive rates and inability to detect mutations below 5% AF, especially in samples with low tumor content, such as those from liquid biopsies.
Innovation Solution
The method involves acquiring both target and reference samples, dividing them into replicates for sequencing, and using statistical tests like Student's t-test or negative binomial distribution to compare sequencing data, reducing false positives and enhancing sensitivity by establishing a background error rate for accurate mutation detection at very low AF.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard sequencing and variant calling methods are used, then the process is simple and cost-effective, but the sensitivity is insufficient to detect mutations below 5% allelic frequency, resulting in high false positive rates
Solution Approach 1:
The patent segments the mutation detection process into multiple independent replicate sequencing runs. Each replicate is analyzed separately, and results are integrated through statistical testing. This segmentation allows the system to achieve high sensitivity (detecting mutations at 0.01% AF) by aggregating signals across replicates while maintaining manageable complexity in each individual run.
Solution Approach 2:
The patent adds a statistical dimension to the analysis by implementing replicate-based statistical tests (t-tests, negative binomial distribution). This transforms the detection problem from a single-threshold decision to a multi-dimensional statistical framework that accounts for background error rates, enabling detection at 0.01% AF with controlled false positive rates.
2Measurement precision
If sequencing depth is increased to improve detection sensitivity, then mutation detection capability improves, but operational costs become prohibitively high
Solution Approach 1:
The patent applies partial action by performing moderate-depth sequencing across multiple replicates rather than extreme-depth sequencing in a single run. Each replicate is sequenced to sufficient depth to detect low-frequency variants, and the combined statistical power of multiple replicates achieves the sensitivity equivalent of much higher single-run depth at lower total cost.
Solution Approach 2:
The patent creates multiple copies (replicates) of the sequencing experiment rather than performing one exhaustive sequencing run. These replicate copies are then analyzed through statistical integration, allowing the system to achieve high sensitivity through replication rather than through prohibitively expensive ultra-deep sequencing.
3Measurement precision
If the detection threshold is lowered to increase sensitivity, then more low-frequency mutations are detected, but false positive rates increase significantly
Solution Approach 1:
The patent implements feedback through replicate-based statistical testing. Each replicate provides information about background error rates and variant frequencies. The statistical framework uses this feedback to distinguish true low-frequency mutations from sequencing errors, enabling detection at 0.01% AF while maintaining controlled false positive rates through p-value thresholds and multiple hypothesis testing corrections.
Solution Approach 2:
The patent performs preliminary sequencing in multiple replicates before making final mutation calls. This preliminary action establishes background error profiles and variant frequency distributions that inform the statistical decision-making process, allowing the system to confidently identify true mutations at very low frequencies while filtering out false positives.
4Ease of operation
If tumor content in samples is low, then liquid biopsy becomes feasible and less invasive, but the ability to detect mutations is compromised due to insufficient tumor-derived DNA
Solution Approach 1:
The patent segments the limited tumor-derived DNA signal across multiple replicate sequencing runs. By distributing the analysis across replicates and using statistical integration, the system maximizes the information extracted from low tumor content samples, enabling detection of mutations at 0.01% AF even when tumor DNA represents a small fraction of total DNA in liquid biopsy samples.
Solution Approach 2:
The patent adds a statistical dimension to compensate for low tumor content. Through replicate-based statistical testing, the system amplifies the signal from limited tumor-derived DNA while suppressing noise from normal DNA and sequencing errors, enabling sensitive mutation detection in liquid biopsies where tumor content is inherently low.
Data Source
AI summary
The invention discloses methods and apparatuses for the detection and diagnostics of genetic alterations/mutations in a target sample, which may be a solid tissue or a bodily fluid. A reference sample is also acquired, and the target and reference samples are replicated into multiple target and reference replicates. The replicates are sequenced, and the sequence data is analyzed based on a statistical test. The statistical test compares the measurements between the target and reference replicates at respective allelic indices. True positive calls are then made based on the results of the statistical testing, and the desired genetic alterations/mutations are identified at the base-pair level. The invention may be used for diagnostics related to cancer, auto-immune disease, organ transplant rejection, genetic fetal abnormalities and pathogens.