BBTools

BBTools provides a suite of multithreaded programs for processing and analyzing DNA and RNA high-throughput sequencing data.

Key Features:

• Multithreading and performance: Implements multithreaded, high-speed processing optimized for large sequencing datasets and high CPU core counts.
• File format support: Processes FASTQ, FASTA, SAM, SCARF, FASTA+QUAL and their compressed or raw variants.
• Automatic detection: Automatically detects quality-score encoding and interleaving of reads.
• BBMerge paired-end merging: Merges paired-end shotgun reads to produce merged sequences for downstream analyses.
• Non-overlapping read merging via k-mers: Merges non-overlapping read pairs using k-mer frequency information to assemble unsequenced gaps between reads.
• Accuracy: Maximizes accuracy in read merging to reduce errors that affect downstream analyses.
• Benchmarking: Has been benchmarked against eight other merging tools using synthetic and real-world datasets for evaluations of speed, accuracy, and scalability.

Scientific Applications:

• Genome assembly: Improves assembly by providing merged reads that span insert regions and reduce redundancy.
• Binning: Supplies higher-quality merged reads that can improve metagenomic binning accuracy.
• Mapping: Produces longer contiguous sequences for read mapping and alignment tasks.
• Annotation: Generates merged sequences that enhance gene prediction and functional annotation.
• Taxonomic analysis and clustering: Provides merged reads that support taxonomic classification and clustering analyses.
• Error-sensitive analyses: Supports analyses where minimizing merging-related errors is critical.

Methodology:

Implements multithreaded processing; automatic detection of quality encoding and interleaving; support for FASTQ, FASTA, SAM, SCARF and FASTA+QUAL (including compressed/raw variants); BBMerge for merging paired-end shotgun reads including non-overlapping pairs using k-mer frequency information; and benchmarking against eight other merging tools on synthetic and real-world datasets.