zDUR: reference-free FASTQ compressor with high compression ratio and speed.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• High-throughput sequencing generates large FASTQ files, requiring efficient compression for storage and transmission.
• Specialized FASTQ compressors offer better performance than general-purpose ones by exploiting data redundancy.
• Existing FASTQ compressors often have limited data applicability or compromise between compression ratio and speed.

Purpose of the Study:

• To develop a reference-free FASTQ compressor with broad data applicability and a balance between compression ratio and speed.
• To address the limitations of existing specialized FASTQ compressors in terms of data compatibility and performance trade-offs.

Main Methods:

• Development of zDUR, a novel reference-free FASTQ compression algorithm.
• Benchmarking zDUR against six other reference-free compressors using 15 diverse datasets.
• Evaluation of compression ratio, speed, and applicability across single-cell RNA-seq, spatial transcriptomics, and other sequencing data types.

Main Results:

• zDUR demonstrates a favorable balance between compression ratio and speed across various sequencing data types.
• Benchmarking on 15 datasets shows zDUR's broad applicability.
• zDUR achieves over a tenfold runtime performance increase on single-cell RNA-seq and spatial transcriptomics data compared to SPRING, while maintaining superior compression ratios.

Conclusions:

• zDUR provides a scalable and efficient solution for reference-free FASTQ data compression.
• The compressor balances performance, speed, and usability for diverse next-generation sequencing datasets.
• zDUR represents an advancement in handling the storage and transmission challenges of large sequencing data volumes.