Genomic Variant Calling via Directed Acyclic Graph Path Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional reassembly-based variant calling methods are computationally expensive and inefficient due to the need to evaluate a large number of candidate sequences, particularly when dealing with large genomes like the human genome, where regions with low homology to the reference sequence complicate alignment and lead to incorrect or unmapped sequence reads.
Innovation Solution
A method that scores variant sites in an assembly graph to reduce the number of candidate sequences evaluated, using a directed acyclic graph (DAG) to rank paths based on sequence read support and exclude low-probability k-mers, allowing for fewer paths to be considered, thereby improving computational efficiency and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If de novo assembly is performed to accurately identify sequences in regions with low homology to reference, then measurement precision is improved, but device complexity and computational resources increase significantly
Solution Approach 1:
The patent segments the genome into smaller regions (e.g., 1kb windows) and performs assembly independently in each region. This divides the computationally expensive de novo assembly task into manageable pieces, reducing memory requirements and enabling parallel processing while maintaining accuracy in low-homology regions
Solution Approach 2:
The patent applies assembly only to specific regions where homology to the reference is low or variable, rather than performing assembly across the entire genome. This localized approach concentrates computational resources where they are most needed, improving accuracy for difficult regions while avoiding unnecessary computation in high-homology regions
2Measurement precision
If local assembly is performed to reassemble sequences from small regions with suspected variation, then measurement precision is improved, but productivity decreases due to significant computational resources required
Solution Approach 1:
The patent performs preliminary alignment of sequence reads to the reference genome before assembly, using efficient alignment algorithms to quickly identify regions of interest. This preliminary step filters out most of the genome, so that expensive assembly is only performed in small regions where variation is suspected, dramatically improving computational efficiency
Solution Approach 2:
The patent performs assembly in a partial manner, only in regions where it is likely to provide value (areas with low homology or suspected variation). This partial assembly approach achieves sufficient accuracy for variant calling without the excessive computational cost of complete de novo assembly
3Measurement precision
If conventional reassembly methods evaluate all candidate sequences to ensure accuracy, then measurement precision is improved, but productivity decreases due to large number of sequences to evaluate
Solution Approach 1:
The patent extracts and evaluates only the most promising candidate sequences based on preliminary scoring criteria (such as read support, consistency with reference in flanking regions, and assembly quality metrics). By taking out only the top candidates for detailed evaluation, the method maintains accuracy while dramatically reducing the number of sequences that require full computational analysis
Data Source
AI summary
In one embodiment, a method for identifying candidate sequences for genotyping a genomic sample comprises obtaining a plurality of sequence reads mapping to a genomic region of interest. The plurality of sequence reads are assembled into a directed acyclic graph (DAG) comprising a plurality of branch sites representing variation present in the set of sequence reads, each branch site comprising two or more branches. A path through the DAG comprises a set of successive branches over two or more branch sites and represents a possible candidate sequence of the genomic sample. One or more paths through the DAG are ranked by calculating scores for one or more branch sites, wherein the calculated score comprises a number of sequence reads that span multiple branch sites in a given path. At least one path is selected as a candidate sequence based at least in part on its rank.