Mutational Burden Evaluation Using Wavelet Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for computing mutational burden are limited to solid tumor samples, require high purity and coverage depths, and are not suitable for liquid biopsy samples, leading to low prognostic efficacy and accuracy, especially in samples with purity under 10%, which is common in liquid biopsies.
Innovation Solution
A method that calculates mutational burden and sample purity for both solid tumor and liquid biopsy samples using allelic depth data, segmenting with wavelet transforms, estimating parameters, classifying segments, and determining burden through intermediate values fitted to a heavy-tail distribution, eliminating the need for germline controls and high coverage depths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current methods are used to compute mutational burden, then accuracy is improved for solid tumor samples, but the method becomes unsuitable for liquid biopsy samples and requires high purity and coverage depths
Solution Approach 1:
The method is designed to universally apply to both solid tumor samples and liquid biopsy samples, eliminating the need for separate analysis protocols. The algorithm can process various sample types (tumor tissue, blood, plasma, urine) and sequencing data formats (panel-based, whole exome, whole genome) through a unified computational framework that adapts to different input characteristics
Solution Approach 2:
The method dynamically adjusts computational parameters based on sample characteristics. It automatically estimates purity and coverage depth metrics, adapts classification thresholds based on observed data distributions, and modifies segmentation parameters to suit different sequencing depths and panel compositions, thereby maintaining accuracy across diverse sample types without requiring manual parameter optimization
2Measurement precision
If current methods are used, then mutational burden can be computed for solid tumors, but they require iterating through purity values which increases computational complexity
Solution Approach 1:
The method performs preliminary estimation of purity and coverage depth metrics before the main mutational burden computation. By pre-calculating these parameters and using them to establish appropriate classification thresholds and segmentation parameters, the algorithm eliminates the need for iterative purification processes, significantly reducing computational complexity while maintaining accuracy
Solution Approach 2:
The method creates simplified representations of the complex data through segmentation into distinct categories (e.g., somatic vs. germline variants, different functional classes). This copying approach allows the algorithm to process data in manageable units rather than handling the entire complex dataset at once, reducing computational burden while preserving essential information for accurate mutational burden calculation
3Measurement precision
If high coverage depths of 250x or greater are required, then accuracy is improved, but the method becomes less suitable for liquid biopsy samples with lower coverage
Solution Approach 1:
The method dynamically adapts its requirements and processing parameters based on the actual coverage depth and data quality of the input sample. It automatically adjusts classification thresholds, segmentation parameters, and statistical significance levels to match the observed data characteristics, enabling accurate mutational burden computation across a wide range of coverage depths from low-coverage liquid biopsies to high-coverage solid tumor samples
Solution Approach 2:
The algorithm changes its operational parameters based on sample type and coverage depth. For low-coverage liquid biopsy samples, it employs different statistical thresholds and segmentation strategies compared to high-coverage solid tumor samples. This parameter adaptation allows the method to maintain accuracy across diverse sequencing depths without requiring a minimum coverage threshold
4Measurement precision
If current algorithms are used, then mutational burden can be computed, but they are panel dependent which reduces versatility
Solution Approach 1:
The method is designed to universally process different panel compositions, gene sets, and sequencing strategies through a unified algorithmic framework. It automatically adapts to the specific genes, regions, or markers being analyzed by detecting data characteristics and adjusting parameters accordingly, making it applicable to cancer panels, somatic panels, germline panels, and other specialized assays without requiring panel-specific algorithms
Data Source
AI summary
The present disclosure relates to a method of evaluating a mutational burden in a sample. The method includes providing first data which represents allelic depths of the sample; calculating rate data based on the first data; segmenting the rate data into a plurality of segments of rate data using a wavelet transform; estimating parameters for each of the plurality of segments; classifying each of the plurality of segments based on the estimated parameters. The classifying includes, for each of a plurality of classification thresholds, generating a classification for the plurality of segments based on the classification threshold and determining an intermediate value based on the classification. The method further comprises determining a mutational burden of the sample based on the intermediate values.