Meraculous

Meraculous assembles de novo genomes from deep paired-end short reads to produce contiguous contigs and scaffolds for genome reconstruction from high-throughput sequencing data such as Illumina.

Key Features:

• Algorithmic Approach: Traverses the k-mer (de Bruijn) graph focusing on subgraphs with unique high-quality extensions and does not rely on an explicit error-correction step.
• Memory-efficient hashing: Uses a memory-efficient hashing scheme to reduce memory footprint for large datasets.
• Assembly performance: Assembled over 95% of the 15.4 megabase haploid yeast Pichia stipitis genome with more than half the sequence in contigs >101 kilobases and scaffolds >269 kilobases and with no errors reported.
• Chromosome-level assembly: Incorporates fosmid ends to enable recovery of entire chromosomes.
• Contig ordering and gap closure: Orders and orients contigs using paired reads with insert sizes of approximately 280 bp or 3.2 kbp and closes many gaps through strategic placement of paired-end reads.

Scientific Applications:

• Microbial genomics: Assembly of microbial genomes exemplified by the high-quality reconstruction of Pichia stipitis.
• Genome research and development: Applicable to broader genome projects, including larger and more complex genomes, where efficient memory use and scaffolding strategies support scalable assembly.

Methodology:

Traverses k-mer (de Bruijn) graphs focusing on subgraphs with unique high-quality extensions; employs a memory-efficient hashing scheme; orders and orients contigs using paired reads (~280 bp and 3.2 kbp); incorporates fosmid ends for chromosome recovery; closes gaps via paired-end read placement; does not perform explicit error correction.