RNA-DNA differences in variant calls from cattle tissues result in erroneous eQTLs

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Genomic variants influence phenotypes, and molecular phenotypes like gene expression are crucial for understanding this link.
• RNA sequencing offers insights into gene expression and splicing, complementing DNA sequencing for variant discovery.

Purpose of the Study:

• To evaluate the utility of RNA sequencing for variant genotyping and its impact on association testing.
• To quantify RNA-DNA differences and their effect on genotype accuracy and eQTL (expression Quantitative Trait Loci) detection.

Main Methods:

• Genotyping of 74 Braunvieh cattle using both DNA and deep total RNA sequencing from multiple tissues.
• Variant calling from RNA sequencing data and comparison with DNA sequencing data.
• Association testing using variants from both data types to identify eGenes and eQTL.

Main Results:

• RNA sequencing identified ~40% of DNA-derived variants with >80% precision, reaching ~98% precision in highly expressed coding regions.
• Allele-specific expression and post-transcriptional modifications were found to reduce variant genotyping accuracy from RNA.
• RNA-derived variants detected ~75% of eGenes found with DNA-derived variants, showing a 2-fold enrichment of eQTL variants, but only 9% of top associated variants were shared.

Conclusions:

• Hundreds of thousands of RNA-DNA differences were observed between variants called from RNA and DNA sequencing.
• Association testing with RNA sequencing variant genotypes identified significant eQTL not found with DNA, but this may lead to increased false positives.
• Caution is advised when analyzing or imputing variants from RNA sequencing data, beyond standard quality control measures.