ChIPWig

ChIPWig compresses ChIP-seq Wig data using lossless and lossy methods to reduce storage while preserving signal features required for downstream analyses such as peak calling (NarrowPeaks) and motif discovery.

Key Features:

• Lossless Compression: Reduces Wig file sizes to approximately 6% of the original without information loss.
• Nonuniform Quantizer Design: Uses the asymptotic theory of optimal point density design for nonuniform quantizers to optimize compression.
• Lossy Compression: Provides a lossy mode that further halves file size relative to lossless compression while preserving peak calling and motif discovery performance with NarrowPeaks methods.
• Random Access: Enables random access to compressed data so specific regions can be retrieved without full decompression.
• Summary Statistics Lookups: Allows retrieval of summary statistics directly from compressed files.
• Performance Efficiency: Compression and decompression average around 0.2 seconds per megabyte on general-purpose computers.
• Format Support: Targets ChIP-seq Wig file format.

Scientific Applications:

• ChIP-seq data storage reduction: Enables large-scale storage reduction for ChIP-seq datasets, including consortium-scale data such as ENCODE.
• Peak calling and motif discovery: Preserves features needed for peak calling and motif discovery analyses, including NarrowPeaks-based methods.

Methodology:

Implements lossless and lossy compression via nonuniform quantization based on the asymptotic theory of optimal point density design for nonuniform quantizers.