lncRNA Genetic Regulation and Disease Mapping
Summary¶
de Goede, Nachun, Ferraro, et al. (2021) profile the expression, tissue specificity, genetic regulation, and disease relevance of 14,100 long non-coding RNA (lncRNA) genes using GTEx v8 multi-tissue transcriptomic and genetic data across 49 tissues, connecting them to 101 complex traits. Most lncRNAs remain uncharacterized despite known regulatory roles for a handful of well-studied examples (HOTAIR in cancer, BACE1-AS in Alzheimer's disease); this study finds eQTLs for 67.3% of annotated lncRNA genes and identifies 1,432 significant lncRNA gene-trait associations, 800 of which are not explained by a stronger effect from a neighboring protein-coding gene.
Tissue Specificity¶
Using a detection method designed to avoid overestimating tissue specificity from simple TPM thresholding (comparing each lncRNA's read count against a length-matched non-genic background region), 316 lncRNA genes were detected in only one broad tissue category — most often testis, brain, blood, and skin. Separately, among lncRNA eGenes (genes with a detected eQTL), 2,783 showed tissue-specific genetic regulation — a higher proportion than seen for protein-coding genes — and this was not purely driven by tissue-restricted expression: 15% of tissue-specific eGenes were nonetheless expressed across all tissue categories (median 24% across tissues, ranging 6% in testis to 100% in uterus), showing that a gene can be broadly expressed while its genetic regulation is tissue-restricted.
Disease and Trait Associations¶
Colocalization analysis identified 1,432 lncRNA gene/splice-cluster colocalization events spanning 69 traits and 166 lncRNA features (119 genes, 47 splice clusters) — versus 9,167 protein-coding colocalization events across 82 traits and 1,096 features in the same framework, giving a sense of relative scale. Trait categories with disproportionately many lncRNA colocalizations included lupus, multiple sclerosis, and blood cell counts; others (ALS, Parkinson's disease, substance use) had none, largely reflecting low GWAS power for those traits rather than an absence of lncRNA involvement. Specific disease associations highlighted included lncRNA QTLs linked to inflammatory bowel disease, type 1 and type 2 diabetes, and coronary artery disease, plus rare-variant associations with BMI.
Significance¶
This establishes a systematic, genome-wide resource for prioritizing which of the thousands of annotated lncRNA genes are plausible contributors to specific complex traits — extending MetaXcan-family colocalization approaches, previously applied mostly to protein-coding genes, to the lncRNA compartment of the transcriptome.
Availability¶
Protected GTEx data: dbGaP (accession phs000424.v8), accessible via the AnVIL platform. Open-access data and visualizations: GTEx Portal.
See Also¶
- Cell-Type Interaction QTLs — a companion GTEx v8 analysis extracting cell-type-specific regulatory signal from the same bulk-tissue data.
- Statistical Colocalization — the methodology used to link lncRNA QTLs to GWAS loci here.