High resolution of full-length RNA sequencing deciphers massive transcriptome complexity during zebrafish embryogenesis.

Background: The zebrafish has significantly advanced our understanding of human disease and development, with nearly 70% of single-copy protein-coding genes conserved between the species. However, research on zebrafish is limited by gaps in existing genome annotations, which are primarily based on computational predictions and short-read sequencing data.

Results: To address this issue, we employed the PacBio Sequel II platform to generate a time-series full-length transcriptome landscape of zebrafish embryogenesis, covering 21 time points from embryo to six days post-fertilization. Our analysis uncovered 2113 previously unannotated genes and 33,018 novel isoforms of previously annotated genes, substantially expanding the current zebrafish gene annotations. We verified these findings using various methods, including domain prediction, homology analysis, conservation analysis, transcript quantification with short-read RNA-seq, and promoter position information with H3K4me3 and CAGE-seq. Furthermore, we analyzed the dynamic expression of transcripts across the 21 developmental stages using next-generation sequencing data, identifying variable splicing events throughout these stages.

Conclusions: Collectively, our study provides a high-resolution and significantly improved transcriptome annotation during zebrafish embryogenesis, offering a valuable resource for the zebrafish research community.