Bandage is a free, open-source Mac application that renders genome assembly graphs as interactive visual networks, letting researchers explore the complex topology of de novo DNA assemblies directly on their desktop.
What is Bandage?
Bandage — short for a Bangin' Assembly with Nodes Designed for Graphical Exploration — is a bioinformatics tool purpose-built for visualising sequence assembly graphs produced by assemblers like SPAdes, Velvet, and Trinity. Where most pipeline output is a flat FASTA file that hides the underlying structure, Bandage exposes the full graph: nodes, edges, dead ends, bubbles, and all the tangled ambiguity that a contig list quietly buries.
I first opened it expecting a niche command-line wrapper with a thin GUI bolted on. What I found was a genuinely usable desktop application — responsive graph rendering, smooth pan-and-zoom, and a search interface that lets you jump straight to a node of interest without scrolling through thousands of edges by hand.
What does Bandage do best?
Bandage excels at turning an opaque assembly graph file (GFA or LastGraph format) into something a human can actually reason about. Load a graph, and within seconds you see which regions assembled cleanly into long linear paths, which collapsed into dense hairballs, and where repetitive elements are forcing the assembler into ambiguous branches.
- Interactive node browsing: click any node to read its sequence, length, coverage depth, and connected neighbours in the info panel.
- BLAST integration: run a BLAST search against the graph directly from the interface — query sequences light up in colour so you can locate a gene of interest inside a messy subgraph in seconds.
- Custom labelling and colouring: colour nodes by depth, by length, by custom CSV annotation, or by BLAST hit — essential when comparing assembly quality across samples.
- Subgraph extraction: select a region of the graph and export just that subgraph or its sequences, bypassing the need to write custom parsing scripts.
- Command-line mode: the same binary accepts flags for headless rendering, so Bandage fits inside automated QC pipelines as well as interactive sessions.
It handles graphs from small bacterial assemblies through to eukaryotic draft genomes — though very large graphs (tens of millions of nodes) will push the renderer hard and reward a machine with ample RAM.
Is Bandage free?
Yes — Bandage is completely free to download and use, including for commercial and institutional research. The source code is published on GitHub under an open-source licence. There is no pro tier, no usage cap, and no telemetry. Installation via Homebrew Cask is a single command away for anyone already using that package manager.
Who should use Bandage?
Bandage is built squarely for computational biologists, genomics researchers, and bioinformaticians who run de novo assembly pipelines and need more than a contig N50 statistic to evaluate quality. If you hand sequence data to SPAdes or Velvet and then wonder what actually happened inside the assembler, Bandage is where you find out.
It is less relevant if your work is purely reference-based (mapping reads to an existing genome) or if you only need summary statistics — tools like Quast or BUSCO cover that ground efficiently. But the moment you want to understand why a particular region assembled poorly, or you suspect a contamination bubble, or you are trying to close a gap manually, Bandage becomes indispensable. Graduate students especially benefit from the visual intuition it builds around graph topology — concepts like dead ends and unresolved repeats click far faster when you can see them.
How does Bandage compare to alternatives?
The honest answer is that direct competitors are rare. Bandage largely created its own category when it launched. Tablet and IGV are excellent for viewing read alignments against a reference, but neither was designed to render assembly graphs. Gfaviz and AGB are newer graph-visualisation tools worth knowing, and they handle large GFA files with some different layout strategies — but Bandage's BLAST integration and its head-start on documentation and community tutorials give it a meaningful practical edge for most lab workflows. If you are working with very large pan-genome graphs, tools in the vg ecosystem may complement rather than replace Bandage.
What are Bandage's limitations?
Very large assembly graphs — think complex eukaryotic genomes assembled with short reads — can produce layouts that take minutes to compute and are difficult to interpret even when rendered. The force-directed layout algorithm is effective for mid-size graphs but can produce visually noisy results at scale. The application has not seen major UI updates in some time, and a handful of rough edges in the preferences system reflect that age. It also requires a working BLAST installation for the sequence-search feature, which adds a setup step on machines that do not already have NCBI BLAST+ configured.
