Variant Review with the Integrative Genomics Viewer

Abstract

Manual review of aligned reads for confirmation and interpretation of variant calls is an important step in many variant calling pipelines for next-generation sequencing (NGS) data. Visual inspection can greatly increase the confidence in calls, reduce the risk of false positives, and help characterize complex events. The Integrative Genomics Viewer (IGV) was one of the first tools to provide NGS data visualization, and it currently provides a rich set of tools for inspection, validation, and interpretation of NGS datasets, as well as other types of genomic data. Here, we present a short overview of IGV's variant review features for both single-nucleotide variants and structural variants, with examples from both cancer and germline datasets. IGV is freely available at https://www.igv.org Cancer Res; 77(21); e31-34. ©2017 AACR.

Figure 1

This figure illustrates five examples of how IGV was used to visually highlight anomalies in aligned sequencing data to aid the investigator in their interpretation. A. Example of a false positive in a GC-rich region. Bases that do not match the reference genome are drawn with the letter of the called base (A, C, G, or T). Sorting and coloring alignments by strand reveal a strong strand bias indicative of a sequencing artifact. Pink and blue reads were aligned to the forward and reverse strand, respectively. B. A germline mutation mischaracterized as a somatic NOTCH2 mutation in a clinical cancer sample. The mutation caller used a mapping quality filter with threshold set to 20, removing most alignments that support the variant from the normal sample and resulting in a reference call. The short horizontal black lines in the alignments represent 2-nucleotide deletions, and the colored bars in the grey alignments represent bases that did not match the reference sequence. In particular, the green bars represent bases that were called as A’s. C. A clinically actionable multi-nucleotide deletion-insertion event (L747_A750delinsP) initially misclassified as two independent events. The horizontal black lines in the grey alignments represent deletions in alignments that otherwise match the reference sequence. The blue column of C’s indicates a large number of mismatches against the G at that locus of the reference sequence. D. Comparison of NGS paired-end short reads (top panel) and third-generation long reads (bottom panel) spanning a 500 base pair inversion. The short reads are colored by pair orientation. The two reads of each pair were expected to point inward toward each other when aligned to the reference sequence. The teal color represents pairs where both reads unexpectedly aligned in the “right” direction, and the dark blue represents pairs where both aligned in the “left” direction. Breakpoints can be inferred from the short reads by careful examination of the outer read extents and dip in coverage. The long reads are split into multiple alignments and each alignment color-coded by strand. Pink and blue segments were aligned to the forward and reverse strand, respectively. The breakpoints are directly visible as the boundaries between alternating strands. E. PacBio long reads support a 4.3 kb translocation from chr19 to chr16 in the SK-BR-3 breast cancer cell line. A black line represents a deletion, and a purple box is an insertion. Deletions and insertions are labeled with the event size.