Methodology for whole genome sequencing of MRSA in a routine hospital

There is growing evidence for the value of bacterial whole genome sequencing in hospital 19 outbreak investigation. Our aim was to develop methods that support efficient and accurate 20 low throughput clinical sequencing of methicillin-resistant Staphylococcus aureus (MRSA). 21 Using a test panel of 25 MRSA isolates associated previously with outbreak investigations, 22 we devised modifications to library preparation that reduced processing time by 1 hour. We 23 determined the maximum number of isolates that could be sequenced per run using an 24 Illumina MiniSeq and a 13 hour (overnight) run time, which equated to 21 MRSA isolates 25 and 3 controls (no template, positive and negative). Repeatability and reproducibility assays 26 based on this sequencing methodology demonstrated 100% accuracy in assigning species 27 and sequence type (ST) and detecting mecA . Established genetic relatedness between isolates 28 was recapitulated. Quality control (QC) metrics were evaluated over nine sequencing runs. 29 168/173 (97%) test panel MRSA genomes passed QC metrics based on the correct species 30 assigned, detection of mecA and ST, and depth/coverage metrics. An evaluation of 31 contamination in these 9 runs showed that positive and negative controls and test MRSA 32 sequence files contained <0.14% and <0.48% of fragments matching another species, 33 respectively. Deliberate contamination experiments confirmed that this was insufficient to 34 impact on data interpretation. These methods support reliable and reproducible clinical 35 MRSA sequencing with a turnaround time (from DNA extraction to availability of data files) 36 of 24 hours. 37