Written and maintained by CASRAI Editorial Board
Last updated
IGV, the Integrative Genomics Viewer, is a free, interactive tool for the visual exploration of genomic data. Researchers use it to look at sequencing alignments, variants and annotation tracks on a reference genome and to check what an analysis pipeline reported. It comes in three forms: a Java desktop application, a browser-based IGV-Web, and the igv.js JavaScript component.
What IGV is used for in research
The IGV project describes the software as a high-performance, easy-to-use, interactive tool for the visual exploration of genomic data. Its documentation focuses on the desktop application, and the software is released under an MIT open-source licence with source code on GitHub (igvteam/igv).
In practice a researcher opens aligned reads, typically a BAM file, over a reference genome and inspects a region by eye: does the read pile-up support a variant call, is coverage even, does a deletion or insertion appear in the reads. The IGV team has published a paper specifically on using IGV for variant review (Robinson et al., Cancer Research 77(21), 2017). For the wider field, see what is bioinformatics and what is genomics.
Which IGV form to use
| Form | What it is (per the IGV documentation) |
|---|---|
| IGV desktop | The Java desktop application; the main focus of the IGV user guide. |
| IGV-Web | The browser-based version of IGV. |
| igv.js | A JavaScript component for embedding an IGV-style viewer in a web page or application. |
Install or access
Download IGV from the official page, igv.org/doc/desktop/DownloadPage (checked 2026-10-29; it listed IGV 2.19.8, with IGV 3.0 Beta on a separate page). Packages are offered for Mac (Apple silicon and Intel, with or without Java), Windows (installers with or without Java), Linux (a zip with Java) and a command-line zip without Java.
- IGV requires Java 21 or greater as of IGV 2.19.1. Packages without Java need Java 21 on your path.
- The IGV Mac apps require macOS 11 (Big Sur) or greater. On older macOS, the download page points to the command-line version or the 2.18.4 Mac app.
- The Linux package with bundled Java covers x64 only; on other architectures use the command-line package with your own Java.
Step-by-step: load a genome and a BAM file
Menu names below follow the IGV desktop user guide pages on reference genomes and loading data (checked 2026-10-29).
- Choose the reference genome. The genome dropdown at the left end of the toolbar shows the current genome and recently loaded ones. For a hosted genome open the Genomes menu and choose Select Hosted Genome, pick the genome and click OK (internet connection required). For your own FASTA or .2bit sequence, an IGV genome JSON file or a GenBank (.gbk) file, use Genomes > Load Genome from File or Load Genome from URL. A sequence file alone loads no gene annotations; a genome JSON file can include them. Switching genomes clears the current session.
- Choose File > Load from File, select your alignment file in the Select Files window and click OK. Keep the index file (for a BAM, the .bai; see BAM index (.bai) files) next to the data file.
- To load from the web instead, choose File > Load from URL, enter the URL and, if the file is indexed, enter the index file name in the field provided, then click OK. The web server must support HTTP byte-range requests for BAM, TDF and other indexed files.
- Zoom in to the region you want to inspect. Alignments are not drawn when the view is wider than the visibility threshold (see Troubleshooting).
- To start over, choose File > New Session (reference genome tracks stay), or right-click a track and choose Remove Tracks.
If the current genome has hosted tracks, a File > Load Hosted Tracks entry appears; tick the datasets and click OK. Ticking a folder selects all its subfolders and datasets.
Output and how to read it
IGV shows its view as stacked tracks. For alignment data, the IGV documentation (viewing alignments basics, checked 2026-10-29) describes these defaults:
| What you see | What it means |
|---|---|
| Coverage track bar chart | Read depth at each position. A bar is grey when all reads match the reference. |
| Coloured coverage bar | Coloured by base counts when a base differs from the reference in more than 20% of quality-weighted reads (the default threshold). Clicking a bar shows per-nucleotide counts and the strand split. |
| Mismatched bases in reads | A green, C blue, G yellow, T red; matching bases are grey. Low base quality (phred) makes a base more transparent. |
| Black bar in a read | A deletion. |
| Purple I | An insertion. |
| Hollow read | Mapping quality of 0. |
| Read colour by strand | Positive strand pink, negative strand purple. |
| Black rectangle under the coverage track | Downsampling happened in that region. The coverage track still reflects all reads. |
Do not misread a downsampled view
By default IGV downsamples alignments using a 50-base window and keeps a maximum of 100 reads per window; third-generation long-read data is not downsampled by default. Because the coverage track is computed from all reads, use it, not the number of visible reads, when judging depth. For file-format background see SAM vs BAM vs CRAM.
Troubleshooting
Why is my alignment track empty or asking me to zoom in?
IGV hides alignments when the view is larger than the visibility threshold and shows a message to zoom in to a smaller region. The documented defaults are 30 kb for most data (for example Illumina DNA), 300 kb for RNA and 1000 kb for third-generation data.
Why will a BAM file from a URL not load?
For BAM, TDF and other indexed files the server hosting the file must support HTTP byte-range requests. If the file is indexed, the index file name must also be entered in the field in the Load from URL dialog.
IGV will not start
Check the Java requirement: IGV 2.19.1 and later need Java 21 or greater, and the Mac apps need macOS 11 or later.
Free alternatives
| Tool | Description and licence (from the project’s own pages) |
|---|---|
| IGV-Web and igv.js | Browser-based and embeddable forms of IGV itself; IGV is under the MIT licence. |
| JBrowse 2 | jbrowse.org describes it as a genome browser that runs on the web, on your desktop, or embedded in your app; released under the Apache License, Version 2.0. |
FAQ
What does IGV stand for?
Integrative Genomics Viewer.
Is IGV free?
Yes. The IGV documentation states it is completely open for anyone to use under an MIT open-source licence.
How do I cite IGV?
The IGV documentation lists: Robinson et al., Nature Biotechnology 29, 24-26 (2011); Thorvaldsdottir et al., Briefings in Bioinformatics 14, 178-192 (2013); Robinson et al., Cancer Research 77(21), 31-34 (2017) for variant review; and Robinson et al., Bioinformatics 39(1) (2023), btac830, for igv.js.
Sources
- IGV desktop overview: https://igv.org/doc/desktop/ (checked 2026-10-29)
- IGV download page: https://igv.org/doc/desktop/DownloadPage/ (checked 2026-10-29)
- IGV user guide, Loading data: https://igv.org/doc/desktop/UserGuide/loading_data/ (checked 2026-10-29)
- IGV user guide, Reference genome: https://igv.org/doc/desktop/UserGuide/reference_genome/ (checked 2026-10-29)
- IGV user guide, Viewing alignments basics: https://igv.org/doc/desktop/UserGuide/tracks/alignments/viewing_alignments_basics/ (checked 2026-10-29)
- JBrowse 2: https://jbrowse.org/jb2/ (checked 2026-10-29)
