Sparse Project VCF: efficient encoding of population genotype matrices.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Variant Call Format (VCF) is the standard for germline genotype representation.
• VCF file sizes increase rapidly with larger cohorts and more discovered rare variants.

Purpose of the Study:

• To introduce Sparse Project VCF (spVCF) as an efficient evolution of VCF.
• To achieve significant size reduction for VCF files with minimal information loss.

Main Methods:

• Developed spVCF using entropy reduction and run-length encoding.
• Ensured spVCF interoperates efficiently with standard VCF, including tabix-based random access.

Main Results:

• Achieved >10× size reduction in VCF files.
• Demonstrated effectiveness on large datasets like DiscovEHR and UK Biobank whole-exome sequencing cohorts.
• Maintained practically minimal information loss.

Conclusions:

• spVCF offers a highly effective solution for managing large genomic datasets.
• The developed method addresses the growing challenge of VCF file size in population sequencing studies.