SVExpress

SVExpress integrates somatic structural variant breakpoint data with gene expression matrices to identify genes whose expression is associated with nearby SVs, supporting analysis of regulatory consequences of genomic rearrangements in cancer.

Key Features:

• Data Integration: Accepts a table of somatic SV breakpoints and a gene-to-sample expression matrix and constructs a gene-to-sample breakpoint matrix for joint analysis.
• Statistical Modeling: Integrates the breakpoint matrix with expression profiles using linear regression modeling to detect expression changes associated with SV breakpoints.
• Association Identification: Systematically catalogs genes that exhibit consistent expression alterations linked to nearby SV breakpoints.
• Mechanistic Insight Support: Enables investigation of mechanisms such as enhancer hijacking and disruption of topologically associated domains (TADs) for top SV–gene associations.
• Compatibility and Scale: Works with output from common SV calling algorithms and has been applied to analyses involving hundreds of cancer sample profiles (e.g., CCLE).
• Cross-dataset Application: Applied to combined whole-genome sequencing and RNA sequencing datasets including the Cancer Cell Line Encyclopedia (CCLE), The Cancer Genome Atlas (TCGA), and Pan-Cancer Analysis of Whole Genomes.
• Implementation: Implemented using Excel VBA / Visual Basic for Applications macros and R scripts.

Scientific Applications:

• Cancer genomics: Identify genes recurrently altered in expression by structural variants across cancer cohorts.
• Mechanism elucidation: Provide candidate SV–gene pairs for studying enhancer hijacking and TAD disruption as mechanisms of gene deregulation.
• Comparative analysis: Compare SV-associated expression changes across datasets (CCLE, TCGA, Pan-Cancer Analysis of Whole Genomes) to identify consistent patterns of deregulation.

Methodology:

Constructs a gene-to-sample breakpoint matrix from somatic SV breakpoint tables and a gene-to-sample expression matrix, then integrates these using linear regression modeling; implemented in Excel VBA/Visual Basic for Applications and R.