GISTIC

GISTIC, short for Genomic Identification of Significant Targets in Cancer, is a statistical method and accompanying software tool used to identify significantly recurrent somatic copy-number alterations in cancer genomes. It aims to distinguish driver alterations that contribute to oncogenesis from passenger alterations that occur by chance by evaluating both how often an alteration occurs across a cohort and how large it is within individual tumors.

Input data for GISTIC come from genome-wide copy-number profiles of tumor samples, which can be produced by array comparative genomic hybridization, SNP arrays, or sequencing-based copy-number analysis. The method calculates a region-level score that integrates the amplitude of copy-number change and its frequency across samples. Through permutation testing, GISTIC estimates a null distribution and assigns q-values to regions, indicating the statistical significance of observed events.

GISTIC reports significant broad events (such as arm-level aneuploidies) and focal events (smaller, localized alterations). The peak region within each significant locus is identified, and candidate genes within these regions are highlighted as potential drivers. Focal, high-amplitude events are especially informative for pinpointing likely cancer genes, whereas broad events may reflect large-scale chromosomal alterations.

Versions of GISTIC have evolved, with GISTIC 2.0 introducing improvements in modeling broad versus focal events, handling of ploidy, and robustness across samples. GISTIC has become a standard tool in cancer genomics for prioritizing regions and genes for functional validation and for integrating copy-number data with expression and mutation analyses.

Limitations include dependence on the quality and resolution of copy-number data, potential sensitivity to sample size, and the need for careful interpretation of broad versus focal alterations.