Long-read transcriptomics of a diverse human cohort reveals widespread ancestry bias in gene annotation

Understanding gene expression diversity across human populations is essential for accurate genome annotation and disease interpretation. However, existing annotations are primarily based on European-derived transcriptomic data, potentially limiting their applicability to other populations. This study aims to assess population-specific transcript diversity and its impact on gene annotation. To achieve this, we performed long-read RNA sequencing on lymphoblastoid cell lines from 43 individuals across eight globally diverse populations. Our workflow included RNA extraction, cDNA synthesis, and sequencing using Oxford Nanopore long-read technology, followed by transcript assembly and comparison with existing gene annotations. We also integrated novel transcripts into reference annotations to evaluate their effect on allele-specific transcript usage detection. This work provides a critical step toward improving transcriptome annotation across diverse populations, ensuring a more comprehensive representation of human genetic variation.