KCMBT

KCMBT counts k-mers in large genomic datasets to enable applications such as genome assembly, repeat detection, multiple sequence alignment, and error detection.

Key Features:

• Novel Trie-Based Algorithm: KCMBT employs a trie-based algorithm that leverages cache-efficient burst tries to store and count k-mers.
• Multi-Threading Capability: KCMBT uses multi-threading to parallelize k-mer counting across multiple CPU cores.
• Performance Superiority: KCMBT is approximately 30% faster than KMC2 and six times faster than Jellyfish2 on human genome datasets, and shows a 20-30% performance advantage over KMC2 across five benchmark datasets under multi-threaded conditions.

Scientific Applications:

• Genome Assembly: Facilitating reconstruction of genomes from short sequencing reads by efficient counting and management of k-mers.
• Repeat Detection: Identifying repetitive sequences within genomes by analyzing k-mer frequencies.
• Multiple Sequence Alignment: Assisting alignment of multiple genomic sequences to identify regions of similarity.
• Error Detection: Detecting sequencing errors via k-mer frequency analysis to support error correction.

Methodology:

The methodology uses a trie-based algorithm with cache-efficient burst tries for k-mer storage and counting, multi-threaded parallelization across CPU cores, and benchmarking against KMC2 and Jellyfish2 on human genome and five benchmark datasets.