Clinical Validation of Genome Reference Consortium Human Build 38 in a Laboratory Utilizing Next-Generation Sequencing Technologies.

Creator(s)

Lisa A. Lansdon, Children's Mercy Kansas CityFollow
Maxime Cadieux-Dion, Children's Mercy Kansas City
John Herriges, Children's Mercy HospitalFollow
Jeffrey J. Johnston, Children's Mercy HospitalFollow
Byunggil Yoo, Children's Mercy HospitalFollow
Joseph Alaimo, Children's Mercy HospitalFollow
Isabelle Thiffault, Children's Mercy HospitalFollow
Neil Miller, Children's Mercy Kansas City
Ana S A Cohen, Children's Mercy HospitalFollow
Elena Repnikova, Children's Mercy HospitalFollow
Lei Zhang, Children's Mercy HospitalFollow
Midhat S. Farooqi, Children's Mercy HospitalFollow
Emily G. Farrow, Children's Mercy HospitalFollow
Carol J. Saunders, Children's Mercy HospitalFollow

Document Type

Article

Publication Date

9-1-2022

Identifier

DOI: 10.1093/clinchem/hvac113

Abstract

BACKGROUND: Laboratories utilizing next-generation sequencing align sequence data to a standardized human reference genome (HRG). Several updated versions, or builds, have been released since the original HRG in 2001, including the Genome Reference Consortium Human Build 38 (GRCh38) in 2013. However, most clinical laboratories still use GRCh37, which was released in 2009. We report our laboratory's clinical validation of GRCh38.

METHODS: Migration to GRCh38 was validated by comparing the coordinates (lifting over) of 9443 internally curated variants from GRCh37 to GRCh38, globally comparing protein coding sequence variants aligned with GRCh37 vs GRCh38 from 917 exomes, assessing genes with known discrepancies, comparing coverage differences, and establishing the analytic sensitivity and specificity of variant detection using Genome in a Bottle data.

RESULTS: Eight discrepancies, due to strand swap or reference base, were observed. Three clinically relevant variants had the GRCh37 alternate allele as the reference allele in GRCh38. A comparison of 88 295 calls between builds identified 8 disease-associated genes with sequence differences: ABO, BNC2, KIZ, NEFL, NR2E3, PTPRQ, SHANK2, and SRD5A2. Discrepancies in coding regions in GRCh37 were resolved in GRCh38.

CONCLUSIONS: There were a small number of clinically significant changes between the 2 genome builds. GRCh38 provided improved detection of nucleotide changes due to the resolution of discrepancies present in GRCh37. Implementation of GRCh38 results in more accurate and consistent reporting.

Journal Title

Clinical chemistry

Volume

68

Issue

9

First Page

1177

Last Page

1183

MeSH Keywords

3-Oxo-5-alpha-Steroid 4-Dehydrogenase; Alleles; Cell Cycle Proteins; Exome; Genome, Human; High-Throughput Nucleotide Sequencing; Humans; Laboratories; Membrane Proteins; Receptor-Like Protein Tyrosine Phosphatases, Class 3

Keywords

GRCh38; clinical laboratory genetics; human reference genome; next-generation sequencing

Recommended Citation

Lansdon LA, Cadieux-Dion M, Herriges JC, et al. Clinical Validation of Genome Reference Consortium Human Build 38 in a Laboratory Utilizing Next-Generation Sequencing Technologies. Clin Chem. 2022;68(9):1177-1183. doi:10.1093/clinchem/hvac113