LiRNA: An Interactive Atlas of Human Liver RNAseq Databases.

Area of Science:

• Genomics
• Bioinformatics
• Translational Medicine

Background:

• Mouse models are crucial for liver disease research, but their translational relevance to human disease is often debated.
• Human transcriptomic data is increasingly used to validate mouse findings, but existing databases are fragmented and difficult to access.
• There is a need for a unified, accessible resource for human liver RNA-sequencing data.

Purpose of the Study:

• To create an interactive atlas of human liver RNA-sequencing datasets to overcome limitations of existing resources.
• To facilitate the validation of translational findings from mouse models to human liver disease.
• To improve the rigor and generalizability of liver disease research.

Main Methods:

• Integrated 17 RNA-sequencing datasets from over 3,000 human liver biopsies into a unified web application (LiRNA).
• Harmonized FASTQ processing, inferred biological sex, and genotyped key metabolic-associated fatty liver disease (MAFLD) variants.
• Systematically evaluated the generalizability of 64 published gene-phenotype and gene-gene correlation findings.

Main Results:

• LiRNA accurately recapitulates known hepatic biology, including fibrosis-associated transcripts and sexually dimorphic gene expression.
• The platform captures genotype-specific transcriptional signatures and response signatures to Hepatitis C virus (HCV).
• Fewer than half of the 64 assessed translational findings generalized consistently across independent, larger cohorts within LiRNA.

Conclusions:

• LiRNA is an open-access, interactive platform integrating harmonized human liver RNA-sequencing data.
• It enables contextualization of translational findings across diverse human populations and disease contexts.
• LiRNA lowers computational barriers, enhancing the rigor and generalizability of liver disease research.