Long-Read Sequencing Reveals RNA Splicing Complexity in Human Diseases.

Transcriptome sequencing is essential for understanding gene expression and RNA features. However, short-read RNA sequencing struggles to analyze complex and full-length messenger RNA molecules. These limitations primarily arise from fragmented read lengths, which make it difficult to accurately characterize alternative splicing patterns, exon structures, or transcription start and termination sites. Long-read RNA sequencing (lrRNA-seq) is an innovative technology that has revolutionized transcriptomic analysis. By end-to-end sequencing, it provides comprehensive insights into transcriptomic structural and regulatory complexity. Moreover, by eliminating the need for transcript assembly and reducing inference errors associated with short-read data, lrRNA-seq can precisely determine exon-intron structures, alternative splicing patterns, transcription initiation and termination sites, alternative polyadenylation, and noncanonical RNA processing events. In this review, we provide a detailed overview of the working principles and technological innovations of lrRNA-seq and emphasize its advantages in transcriptome research. In addition, we systematically assess the methodological aspects, focusing on isoform analysis, quantification, error correction, and algorithm development, which improve the reliability of lrRNA-seq analyses. We further discuss recent applications and developments of lrRNA-seq related to various diseases. Recent studies have revealed disease-related splicing dysregulation, discovered novel pathogenic isoforms, and clarified RNA-mediated mechanisms. Furthermore, we discuss emerging efforts to integrate long-read sequencing with single-cell and spatial transcriptomics, thereby permitting the characterization of splicing complexity across specific cells, tissues, and microenvironments within the whole organism. In conclusion, lrRNA-seq is a transformative technology for advancing disease diagnostics and precision medicine.