Compact NGS Read Formatting via Quality Sub-Sequence Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Next-generation sequencing (NGS) systems face challenges in efficient storage and processing of large genome datasets due to high storage costs and computational demands, primarily because of the large size of raw reads files that include base quality scores, which are not effectively reduced by existing formats like FASTQ without losing critical information.
Innovation Solution
A method and apparatus for formatting raw read data by identifying and representing only the longest sub-sequences of bases with quality scores above a specified threshold, allowing for substantial reduction in file size while retaining essential information, and enabling faster alignment and processing, compatible with existing downstream processing components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If FASTQ format is used to store raw reads with base quality scores, then read quality information is preserved, but storage size increases substantially
Solution Approach 1:
The patent extracts only the essential quality information by identifying and storing the longest sub-sequences of bases with quality scores above a specified threshold, rather than storing all base quality scores. This selective extraction reduces storage requirements while preserving the most critical quality data needed for downstream analysis.
Solution Approach 2:
The patent applies local quality control by differentiating between regions of high quality and low quality within each read. Instead of treating all bases uniformly, it identifies specific sub-sequences that meet quality thresholds, allowing downstream processing to focus on high-quality regions while reducing storage for low-quality portions.
2Measurement precision
If complete base quality scores are stored for all bases, then accurate quality assessment is enabled, but computational processing time increases
Solution Approach 1:
The patent extracts only the essential quality information by identifying and storing the longest sub-sequences of bases with quality scores above a specified threshold, rather than storing all base quality scores. This selective extraction reduces storage requirements while preserving the most critical quality data needed for downstream analysis.
Solution Approach 2:
The patent applies partial action by storing only the portion of quality information that is sufficient for most downstream applications—the longest high-quality sub-sequences—rather than complete quality scores for all bases. This partial representation accelerates processing while maintaining adequate quality assessment capability.
3Reliability
If all reads are processed and stored in detail, then comprehensive genomic coverage is achieved, but storage and computational costs increase
Solution Approach 1:
The patent applies local quality control by differentiating between regions of high quality and low quality within each read. Instead of treating all bases uniformly, it identifies specific sub-sequences that meet quality thresholds, allowing downstream processing to focus on high-quality regions while reducing storage for low-quality portions.
Solution Approach 2:
The patent discards redundant or low-value data by not storing complete quality scores for all bases, particularly for bases that fall below quality thresholds. The essential information is recovered through the sub-sequence representation, which captures the most important quality patterns needed for reliable genomic analysis.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In a method operative on a genetic sequencing read comprising a base sequence acquired by processing a tissue sample, a compact text representation of the genetic sequencing read is generated. The compact text representation includes (1) a text string representing the base sequence and (2) a base quality text field identifying the longest sub-sequence of the base sequence for which base quality scores of the bases of the sub sequence satisfy a base quality score threshold; and storing the compact text representation of the genetic sequencing read in a raw reads storage. To provide flexibility, the base quality text field may identify the longest sub-sequence for each of two or more different base quality score thresholds. During reads alignment, offset boundaries for the genetic sequencing reads can be efficiently chosen using the content of the base quality text field.