.png){kind=icon}
.png){kind=icon}
Return
Invas: an inversion-aware method for transcriptome assembly
.png){kind=icon}
OA
AI
DOI：10.1038/s41467-026-75286-0.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Intragenic inversions reverse sequence orientation within genes and create non-collinear splice junctions that standard transcript assemblers miss, resulting in incomplete reconstruction and biased quantification. Here we show Invas, an inversion-aware framework operating with bulk short-read sequencing data. Invas integrates whole-genome sequencing breakpoints with transcriptomic sequencing evidence, rescues unmapped reads, and assembles isoforms using conjugate flow optimization. In silico across 42,000 events, Invas demonstrates high precision and recall and improves quantification of unaffected transcripts compared with conventional tools. We apply Invas to seven disease cohorts, identifying recurrent germline susceptibility variants and somatic drivers. Furthermore, Invas facilitates the discovery of somatic inversion-derived tumor-specific antigens with strong predicted immunogenicity. We release Invas and InvasDB as community resources to enable accurate characterization of inversion-affected genes, filling a critical gap in structural variant-aware transcriptomics. The authors present Invas, a tool that assembles transcripts affected by intragenic inversions. Applied across multiple cancers, it uncovers recurrent inversions linked to disease progression and identifies inversion-derived immunotherapy targets.
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
15.7
Papers:
9.3W
Citations:
91.2W