LMAS

LMAS evaluates de novo assembly methods for metagenomic sequence data to benchmark assembler accuracy, computational resource usage, and strain-resolution capabilities.

Key Features:

• Flexible Evaluation Platform: Assesses assembler performance using known standard communities to compare outcomes across diverse assemblers and sample types.
• Performance Comparison: Systematically compares assemblers and reports that k-mer De Bruijn graph assemblers generally achieve higher accuracy while imposing increased computational demands.
• Critical Insights on Assembler Performance: Reports that assemblers marketed specifically for metagenomics do not consistently outperform general genomic assemblers and identifies ABySS, MetaHipmer2, minia, and VelvetOptimiser as relatively underperforming on complex samples.
• Strain Resolution Analysis: Documents that achieving meaningful strain resolution at the single-nucleotide polymorphism (SNP) level remains challenging with current assembler technologies.

Scientific Applications:

• Metagenomic assembly benchmarking and assembler selection: Guides researchers in choosing assemblers and planning analyses by informing decisions on computational resource allocation, replicon focus, and study-specific objectives.

Methodology:

LMAS tests various de novo assembly methods against known standard communities and systematically compares assembler performance metrics.