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

Ianthe A E M van Belzen, Casey Cai, Marc van Tuil, Shashi Badloe, Eric Strengman, Alex Janse, Eugène T P Verwiel, Douwe F M van der Leest, Lennart Kester, Jan J Molenaar, Jules Meijerink, Jarno Drost, Weng Chuan Peng, Hindrik H D Kerstens, Bastiaan B J Tops, Frank C P Holstege, Patrick Kemmeren, Jayne Y Hehir-Kwa

Research output: Contribution to journalArticlepeer-review


BACKGROUND: Gene 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.

METHODS: We 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 128 patients by WGS and RNA sequencing.

RESULTS: In a pediatric pan-cancer cohort of 128 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.

CONCLUSIONS: Our 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.

Original languageEnglish
Pages (from-to)618
JournalBMC Cancer
Issue number1
Publication statusPublished - 3 Jul 2023


  • Child
  • Humans
  • Neoplasms/genetics
  • RNA-Seq
  • High-Throughput Nucleotide Sequencing/methods
  • Precision Medicine
  • Sequence Analysis, RNA/methods
  • Gene Fusion
  • Whole Genome Sequencing


Dive into the research topics of 'Systematic discovery of gene fusions in pediatric cancer by integrating RNA-seq and WGS'. Together they form a unique fingerprint.

Cite this