Key Points

Introduction to IGV

A genome browser lets you visually explore genomic data in the context of a reference genome, which helps you spot problems and build intuition that raw files alone do not give you.
IGV is a free, desktop genome browser that runs locally and supports many genomics file formats, including FASTA, BED, BAM, TDF, and VCF.
IGV can display many types of data as tracks, including epigenomic signal, aligned sequencing reads, RNA-seq coverage/junctions, and variant genotypes.

Load a reference genome first, then load data tracks with File > Load from Server… or File > Load from File…; use Genomes > More… to add other hosted genomes beyond the default.
IGV supports four broad zoom levels: whole genome, chromosome, region, and base-pair resolution; drag on the ruler, double-click a track, or use the search box to move between them.
The sequence track shows the reference genome and, optionally, a 3-frame translation; the gene track shows exons, introns, and UTRs, and can be displayed collapsed, expanded, or squished.
Tools > Combine Data Tracks lets you add, subtract, multiply, or divide two numeric tracks into a new derived track.

Load a custom genome sequence with Genomes > Load Genome from File…; load additional annotation tracks (genes, features) with File > Load from File….
A genome loaded from a bare FASTA file has no gene names or cytoband ideogram until you add that information yourself as separate tracks.

Aligned reads are stored in SAM/BAM format; recognizing terms like insert size, mapping quality, and CIGAR string helps you interpret what IGV shows you.
Load a BAM file with File > Load from File…; IGV finds its .bai index automatically as long as it is named correctly and kept alongside the BAM file.
Sorting and coloring aligned reads (by base, by base quality shading, or by read strand) helps you distinguish a real variant from a sequencing or alignment artifact.
Warning signs of a false-positive variant include low base quality and a mismatch that only appears on reads from one strand when it should not.
Tools > BLAT… (or right-click a read and choose BLAT read sequence) checks whether a read aligns uniquely or is a multi-mapping artifact from a repetitive region.

IGV computes BAM coverage on the fly at high zoom levels, but does not attempt this at low zoom levels for performance reasons.
Use Tools > Run igvtools… > Count to precompute a .tdf coverage summary file for a BAM file.
Attach a precomputed .tdf file to a coverage track via right-click > Load pre-computed coverage data….

RNA-seq reads come from spliced mRNA (introns already removed), so a read spanning two exons aligns back to the intron-containing genome as a gapped, “split” alignment.
Enable the splice junction track (via View > Preferences > Alignments) to see, as arcs, which exons are spliced together in RNA-seq data.
Compare coverage and junction tracks across samples or tissues to spot alternative splicing.
A Sashimi plot combines exon coverage (peaks) and junction reads (arcs) in one view, and read-count filtering helps focus on well-supported junctions.

A VCF track in IGV has a variant sites panel and a per-sample genotypes panel; a sample information color bar (driven by an auxiliary tab-delimited file) can show sample metadata such as population.
Variant sites are colored blue (reference allele) / red (alternate allele); genotypes are colored grey (homozygous reference) / blue (heterozygous) / cyan (homozygous alternate).
Search for a variant using gene:protein-change syntax (e.g. APOL1:S342G) as well as by coordinate or gene name.
Grouping genotypes by a sample attribute (e.g. population) makes it easy to compare variant patterns across groups of samples.

IGV can color read pairs by inferred insert size or by orientation, which reveals deletions, insertions, inter-chromosomal fusions, and inversions without running a dedicated SV caller.
A deletion shows up as pairs with a larger-than-expected insert size (colored red) and a coverage drop between two breakpoints.
A fusion shows up as reads colored by an unexpected mate chromosome, clustered at the fusion junction.
An inversion shows up as same-strand (forward-forward or reverse-reverse) discordant pairs at its two breakpoints; right-click View as pairs to see this clearly.

Long reads (PacBio, Oxford Nanopore) trade a higher per-read error rate for much greater length than short Illumina reads.
IGV’s quick consensus mode and indel-hiding thresholds (View > Preferences > Alignments) filter raw long-read noise so consensus signal is easier to see.
A read’s MAPQ score (0–255) reflects alignment confidence; low MAPQ is a warning sign for any variant called from that read.
File > Load from URL… streams a remote BAM/CRAM directly from a server (finding its index automatically), without downloading the whole file first.
Link supplementary alignments connects a single long read’s split pieces on one row; combined with Color alignments by > Read strand (pink = forward, purple = reverse), this reveals an inversion directly within individual reads.
A single long read can directly reveal a structural variant by splitting across its breakpoint, without needing paired-end evidence.
Coloring and grouping reads by haplotype (HP) tag separates two alleles for direct comparison, useful for phased variants and repeat expansions — and for telling a heterozygous structural variant (one haplotype) apart from a homozygous one (both haplotypes).

The core IGV workflow — load a genome and tracks, navigate to a region of interest, then use track-specific sort/color/display options — applies broadly across data types.
IGV has many features beyond this lesson, including a web-embeddable sibling (igv.js) and batch scripting.