Hobbes

Hobbes maps next-generation sequencing (NGS) DNA reads and identifies all possible mapping locations to improve accuracy in applications such as ChIP-seq and RNA-seq.

Key Features:

• Comprehensive Mapping: Unlike Bowtie and BWA that prioritize top candidate mappings, returns all potential mapping positions for each read to enable detection in repeat regions.
• Efficiency and Speed: Leverages additional prefix q-grams for enhanced filtering during read mapping, achieving up to an order of magnitude speedup over state-of-the-art mappers while maintaining similar accuracy.
• Resource Optimization: Memory-efficient implementation that consumes less space compared to competing tools, suitable for large-scale genomic studies.
• Versatile Read Handling: Supports short and long reads, single-end and paired-end configurations, and multithreading across multiple CPU cores.
• Algorithmic Precision: Handles sequence variation using Hamming distance for substitutions and edit distance for substitutions, insertions, and deletions.
• Open Source Availability: Implemented in C++ with source code freely available.

Scientific Applications:

• ChIP-seq binding-site discovery: Enables identification of binding sites within repeat regions by reporting all possible read mappings.
• RNA-seq transcript abundance estimation: Improves transcript abundance estimates by providing comprehensive mapping of reads across transcripts.
• Large-scale NGS projects: Suited for high-throughput studies that require memory-efficient, multithreaded mapping of short and long reads.

Methodology:

Uses additional prefix q-grams for read-filtering, supports Hamming and edit distance alignment models, returns all candidate mappings, and is implemented in C++ with multithreading support.