FLIBase: a comprehensive repository of full-length isoforms across human cancers and tissues

Abstract

Regulatory processes at the RNA transcript le v el pla y a crucial role in generating transcriptome diversity and proteome composition in human cells, impacting both ph y siological and pathological states. This study introduces FLIBase ( www.FLIBase.org ), a specialized database that focuses on annotating full-length isoforms using long-read sequencing techniques. We collected and integrated long-read (351 samples) and short-read (12 469 samples) RNA sequencing data from diverse normal and cancerous human tissues and cells. The current version of FLIBase comprises a total of 983 789 full-length spliced isoforms, identified through long-read sequences and verified using short-read e x on-e x on splice junctions. Of these, 188 248 isoforms have been annotated, while 795 541 isoforms remain unannotated. By overcoming the limitations of short-read RNA sequencing methods, FLIB ase pro vides an accurate and comprehensiv e representation of full-length transcripts. T hese comprehensiv e annotations empo w er researchers to undertak e v arious do wnstream analy ses and in v estigations. Importantly, FLIB ase e xhibits a significant advantage in identifying a substantial number of previously unannotated isoforms and tumor-specific RNA transcripts. These tumor-specific RNA transcripts ha v e the potential to serv e as a source of immunogenic recurrent neoantigens. T his remarkable disco v ery holds tremendous promise f or adv ancing the de v elopment of tailored RNA-based diagnostic and therapeutic strategies f or v arious types of human cancer.