Genetic Sequencing Data Compression Using Homology-Based Read Reordering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for compressing genetic sequencing data, particularly in the FASTQ format, lack support for variable length reads, scalability to high coverage datasets, pairing-preserving compression, and lossless compression, resulting in inefficient data storage and transfer.
Innovation Solution
The method involves reordering and encoding genetic sequencing data by aligning sequencing reads based on homology, generating a reference sequence, and reordering data streams to create a series of blocks for compression, while preserving read order and supporting paired-end sequencing data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If general-purpose compressors are used on FASTQ datasets, then compression is performed, but the inherent redundancy in the structured data cannot be fully exploited
Solution Approach 1:
The patent changes the fundamental parameters of how sequencing data is organized and processed. It transforms the data from a collection of independent variable-length reads into a structured format with a consensus sequence and aligned reads, enabling efficient compression while maintaining support for variable length reads through the alignment framework
2Loss of substance
If existing FASTQ compressors are used, then some compression is achieved, but scalability to high coverage datasets is limited
Solution Approach 1:
The patent segments the compression process into distinct phases: building a consensus sequence from initial reads, then aligning subsequent reads to this consensus. This segmentation enables scalable processing of high coverage datasets by repeatedly applying the same compression logic to blocks of reads rather than attempting to process all reads simultaneously
Solution Approach 2:
The patent performs preliminary action by first constructing a consensus sequence from a subset of reads before compressing the remaining reads. This preliminary consensus building creates a reference framework that dramatically simplifies the compression of subsequent reads, enabling scalability to high coverage datasets
3Loss of substance
If compression methods are applied to paired-end sequencing data, then data size is reduced, but pairing information may be lost
Solution Approach 1:
The patent merges paired-end reads that align to the same region of the consensus sequence into a single compressed representation. By combining information from both reads of a pair while preserving their relationship in the alignment structure, the method achieves compression without losing pairing information
4Loss of substance
If existing compression approaches are used, then some data reduction is achieved, but lossless compression with perfect reconstruction is not possible
Solution Approach 1:
The patent creates a compressed representation that copies essential information from the original reads through alignment to the consensus sequence. Each read is represented by its position and alignment information relative to the consensus, enabling perfect reconstruction of the original reads without loss of information
Data Source
AI summary
Embodiments of the invention are generally directed to compressing genetic sequencing data. In many embodiments, the genetic sequencing data is reordered and encoded based on sequence homology between individual sequencing reads within the genetic sequencing data. Several embodiments are directed to systems to compress genetic sequencing data, and some embodiments are directed to non-transitory, machine-readable media that direct a processor to compress genetic sequencing data. In further embodiments, the genetic sequencing data represents paired-end sequencing data, and several embodiments transmit the data to a remote device.