Whole-genome sequencing for antimicrobial surveillance: species-specific quality thresholds and data evaluation from the network of the European Union Reference Laboratory for Antimicrobial Resistance genomic proficiency tests of 2021 and 2022

作者：Lauge Holm Sørensen, Susanne Karlsmose Pedersen, Jacob Dyring Jensen, Niamh Lacy-Roberts, Athina Andrea, Michael S. M. Brouwer, Kees Veldman, Yan Lou, Maria Hoffmann, René S. Hendriksen · 发表于：mSystems · 年份：2024 · DOI：10.1128/msystems.00160-24 · 被引用次数：12 · 研究领域：Antibiotic Resistance in Bacteria、Bacterial Identification and Susceptibility Testing、Pharmaceutical and Antibiotic Environmental Impacts

ABSTRACT As antimicrobial resistance (AMR) surveillance shifts to genomics, ensuring the quality of whole-genome sequencing (WGS) data produced across laboratories is critical. Participation in genomic proficiency tests (GPTs) not only increases individual laboratories' WGS capacity but also provides a unique opportunity to improve species-specific thresholds for WGS quality control (QC) by repeated resequencing of distinct isolates. Here, we present the results of the EU Reference Laboratory for Antimicrobial Resistance (EURL-AR) network GPTs of 2021 and 2022, which included 25 EU national reference laboratories (NLRs). A total of 392 genomes from 12 AMR-bacteria were evaluated based on WGS QC metrics. Two percent ( n = 9) of the data were excluded, due to contamination, and 11% ( n = 41) of the remaining genomes were identified as outliers in at least one QC metric and excluded from computation of the adjusted QC thresholds (AQT). Two QC metric correlation groups were identified through linear regression. Eight percent ( n = 28) of the submitted genomes, from 11 laboratories, failed one or more of the AQTs. However, only three laboratories (12%) were identified as underperformers, failing across AQTs for uncorrelated QC metrics in at least two genomes. Finally, new species-specific thresholds for “N50” and “number of contigs > 200 bp” are presented for guidance in routine laboratory QC. The continued participation of NRLs in GPTs will reveal WGS workflow flaws and improv...