Systematic discovery of gene fusions in pediatric cancer by integrating RNA-seq and WGS

Abstract

BackgroundGene fusions are important cancer drivers in pediatric cancer and their accurate detection is essential for diagnosis and treatment. Clinical decision-making requires high confidence and precision of detection. Recent developments show RNA sequencing (RNA-seq) is promising for genome-wide detection of fusion products but hindered by many false positives that require extensive manual curation and impede discovery of pathogenic fusions. MethodsWe developed Fusion-sq to overcome existing disadvantages of detecting gene fusions. Fusion-sq integrates and “fuses” evidence from RNA-seq and whole genome sequencing (WGS) using intron-exon gene structure to identify tumor-specific protein coding gene fusions. Fusion-sq was then applied to the data generated from a pediatric pan-cancer cohort of 129 patients by WGS and RNA sequencing. ResultsIn a pediatric pan-cancer cohort of 129 patients, we identified 155 high confidence tumor-specific gene fusions and their underlying structural variants (SVs). This includes all clinically relevant fusions known to be present in this cohort (30 patients). Fusion-sq distinguishes healthy-occurring from tumor-specific fusions and resolves fusions in amplified regions and copy number unstable genomes. A high gene fusion burden is associated with copy number instability. We identified 27 potentially pathogenic fusions involving oncogenes or tumor-suppressor genes characterized by underlying SVs, in some cases leading to expression changes indicative of activating or disruptive effects. ConclusionsOur results indicate how clinically relevant and potentially pathogenic gene fusions can be identified and their functional effects investigated by combining WGS and RNA-seq. Integrating RNA fusion predictions with underlying SVs advances fusion detection beyond extensive manual filtering. Taken together, we developed a method for identifying candidate gene fusions that is suitable for precision oncology applications. Our method provides multi-omics evidence for assessing the pathogenicity of tumor-specific gene fusions for future clinical decision making.

Molecular analysis of cancer genomes in children with Lynch syndrome: Exploring causal associations

“…Exome and/or genome sequencing was performed on a total of 20 tumor and 18 normal samples, according to different procedures depending on the time of sequencing and the origin of the samples (Table S1). Samples from the Princess Máxima Center were handled as described by van Belzen et al 16 For exome sequencing, libraries were captured with KAPA MedExome or KAPA HyperExome (Roche). INFORM registry samples were handled as described by Worst et al 17 For a subset of exome sequencing samples, libraries were captured with Agilent SureSelect version 7.…”