Biomedical Data Analysis Platform

Home > Analysis Platform > Contig binning

Introduction：

MEGAHIT is an ultra-fast, memory-efficient de novo assembler for metagenomic shotgun sequencing data. It uses a succinct de Bruijn graph and iterative k-mer strategies to assemble complex microbial communities on standard hardware with minimal RAM. Note: MEGAHIT performs assembly only; it does not perform contig binning. Binning requires downstream tools (e.g., MetaBAT2, MaxBin2) using MEGAHIT’s output contigs.

Workflow Case Analysis Versions Manual Download

K-mer Counting: Builds compact de Bruijn graph from reads using multiple k-mer sizes.
Graph Simplification: Removes tips, bubbles, and low-coverage edges iteratively.
Contig Extraction: Traverses simplified graph to generate final contigs.
Output: Writes FASTA contigs and assembly statistics.

Parameter	Description
`-1`, `-2`	Paired-end FASTQ files (gzipped OK).
`-r`	Single-end or interleaved FASTQ.
`-o`	Output directory.
`-t`	Threads (default: all CPUs).
`-m`	Memory limit in GB (e.g., `0.9` = 90% of system RAM).
`-k-list`	Custom k-mer list (e.g., `21,33,55,77,99,127`). Default auto-selected.
`--min-contig-len`	Minimum contig length to output (default: 200).
`--presets`	`meta-sensitive` (default), `meta-large`, or `sensitive`. Adjusts k-mer strategy.
`--clean-data`	Remove intermediate files after completion.

Results

Versions

Usage

Core Parameters