Lossless and reference-free compression of FASTQ/A files using GeneSqueeze

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• High-throughput sequencing generates massive data files, demanding efficient storage solutions for omics analytics.
• Longitudinal studies exacerbate storage challenges, hindering the full potential of sequencing technologies in biomedical research.
• Current compression methods may not adequately address the specific needs of sequencing data, impacting storage and transmission costs.

Purpose of the Study:

• To introduce GeneSqueeze, a novel lossless, reference-free compression algorithm for FASTQ/A files.
• To evaluate GeneSqueeze's compression performance and data integrity compared to existing methods like gzip and SPRING.
• To address the critical need for efficient storage solutions in omics analytics.

Main Methods:

• Developed GeneSqueeze, a Python-based, reference-free compression algorithm for FASTQ/A files.
• Implemented an auto-tuning compression protocol based on file distribution.
• Compared GeneSqueeze against gzip and SPRING using metrics such as compression ratio and data loss.

Main Results:

• GeneSqueeze achieved up to three times higher compression ratios than gzip, with comparable ratios to SPRING.
• Both GeneSqueeze and gzip ensured 100% lossless compression across all FASTQ file components.
• SPRING's modes exhibited data loss of non-ACGTN nucleotides and metadata, unlike GeneSqueeze and gzip.

Conclusions:

• GeneSqueeze is a competitive, specialized compression method for nucleotide sequence files.
• The algorithm significantly reduces storage and transmission costs for large omics datasets.
• GeneSqueeze preserves data integrity, making it valuable for biomedical research and personalized patient care.