Exome variant discrepancies due to reference-genome differences | JoVE Visualize

Area of Science:

Genomics and Bioinformatics
Human Genetics
Medical Genomics

Background:

The GRCh37 human reference genome is still widely used despite the release of GRCh38 over seven years ago.
The impact of using different human reference genome assemblies on variant identification in large-scale exome sequencing data for disease association studies has not been quantified.

Purpose of the Study:

To quantify the impact of utilizing GRCh37 versus GRCh38 human reference genomes on identifying single-nucleotide variants (SNVs) and insertion-deletions (indels).
To assess how reference assembly choice affects variant interpretation for Mendelian and common diseases using exome sequencing data.

Main Methods:

Variant calling of SNVs and indels was performed on 1,572 exomes using both GRCh37 and GRCh38 human reference genomes.
Discordant variants and their genomic locations were analyzed, identifying discrete discordant reference patches (DISCREPs).
Genes enriched for discordant variants were identified and cross-referenced with known Mendelian diseases and common phenotypes from genome-wide association studies.

Main Results:

A total of 1.5% of SNVs and 2.0% of indels were discordant between GRCh37 and GRCh38 reference assemblies.
76.6% of discordant variants were located in discrete discordant reference patches (DISCREPs), which constitute only 0.9% of exome sequencing loci.
206 genes showed significant enrichment for discordant variants, with 8 implicated in Mendelian diseases and 53 associated with common phenotypes; reference choice also impacted variant interpretation after lift-over.

Conclusions:

The choice of human reference genome assembly (GRCh37 vs. GRCh38) significantly affects variant identification and interpretation in exome sequencing data.
Discordant variants are concentrated in specific genomic regions (DISCREPs), highlighting potential challenges in variant calling accuracy.
Careful evaluation of genes and loci affected by reference assembly choice is crucial for both research and clinical applications in genetic disease studies.