Presentation Information
[OE5-1]eQTL analysis for full length transcripts using long-read technology reveals a lot of splicing variant-specific eQTLs
○Yuya Nagura, Mihoko Shimada, Akihiro Fujimoto (Department of Human Genetics, Graduate School of Medicine, The University of Tokyo)
Expression quantitative trait loci (eQTL) are genetic variants that are related to variations in gene expression levels. Integration of eQTL and results of genome-wide association studies (GWAS) can identify causal genetic variation of diseases, and many eQTL studies have been conducted. However, most previous eQTL studies used short-read sequencers to examine gene expression levels and did not analyze full-length transcripts. Recently long-read sequencing technologies have become available for RNA-seq, which can generate reads longer than 10 kbp and capture splicing variants (SpVs) accurately. In this study, we conducted Nanopore long-read RNA-seq using B cells of 67 Japanese analyzed in the 1000 Genomes Project. eQTLs at SpV-level and gene-level (sum of SpV in each gene) were identified using the result of RNA-seq and genetic variations (SNP and short indel) in autosomes. We detected the expression of 43,691 SpVs from 15,392 genes, in which 6,834 SpVs were novel. From eQTL analysis at SpV-level, we identified 33,955 eQTLs. Of these, 23,434 were SpV-specific and not detected as gene-level eQTLs. Among SpV-specific eQTLs, 379 overlapped with SNPs in the GWAS catalog database. SpV-specific eQTLs were associated with minor SpVs in each gene (p＜＜0.01). We also found that SpV-specific eQTLs were significantly enriched in H3K36me3 regions compared to all eQTLs (gene-level and SpV-level) (OR=1.46, p=2.7×10-31). This study revealed that a lot of variants can affect the expression level of SpV and that analysis of full-length transcripts identifies a large number of novel eQTLs.