Practical quality control for whole genome sequencing in clinical microbiology John WA Rossen, PhD, MMM Department of Medical Microbiology, University of Groningen, UMCG, Groningen, The Netherlands
Disclosure of speaker s interests (Potential) conflict of interest None Potentially relevant company relationships in connection with event Scientific Collaborations with Checkpoints, BioVisible, Ridom, Bioclear, Qiagen CLC bio no personal benefits Sponsorship or research funding Several National and European Grants
Groningen Sunny Harbor Northern Sun
NGS used to answer patients questions 1. Did you prevent colonization and infection today? à Typing to determine genetic relatedness (isolate) à Tailor-made screenings assays (clinical sample) à Metagenomics (RUO, clinical sample) 2. Do I have an infection/id and which one? à ID of bacteria 16S / 16-23S (clinical sample) à Metagenomics (RUO, clinical sample) 3. What might be the optimal therapy? à Detection of resistance and virulence genes (isolate) à Metagenomics (RUO, clinical sample)
Next generation sequencing in the clinical lab Ion Torrent MiSeq 1 MiSeq 2 Nextseq 3x MinIon Where: @ the Microbiology lab not in our core facility Why: Need for Speed Infectious samples Customized protocols Building experience and capacity Gaining knowledge Lean: improved workflow
Lab design and QC Erwin RaaNGS e-lab wet-lab Sigrid Rosema
Validation of WGS (ISO 15189) Reproducibility Repeatability Comparison with established methods For all bacteria?
Validation WGS in UMCG Three marker bacteria MRSA VRE Klebsiella pneumoniae WGS used for Typing Resistome Virulome
Reproducibility and repeatability Wet-lab One MRSA, one VRE and one Klebsiella pneumonia strain Two technicians two different days Three times DNA isolation, three times library prep Three times MiSeq E-lab Each sequence three times assembled (CLC Genomics Workbench) Each assembly Typed using a gene-by-gene approach (cgmlst Seqsphere) Resistome - Resfinder (Center for Genomic Epidemiology) Virulome Virulencefinder (Center for Genomic Epidemiology)
NGS and MLST core genome MLST à nomenclature (CT) accessory genome MLST cgmlst or gene-by-gene typing after sequencing whole genome A B C D
Results - typing MRSA VRE K. pneumoniae
K. pneumoniae - Comparison with established methods Tabel 7: Overzicht Klebsiella pneumoniae stammen Externe nr. Monsternr. RIVM-nr. AFLP Type (VUMC) MLST (UMCG) Type Code Diversilab (UMCG) Stam 9 189591 V 1181200202 AT12187 336 2-1-1-1-72-4-4 (1) Stam 14 192536 W 1181200203 AT12188 307 4-1-2-52-1-1-7 (2) Stam 20 179690 I 1181200204 AT12189 15 1-1-1-1-1-1-1 (3) Stam 23 188050 C 1181200205 AT12189 15 1-1-1-1-1-1-1 (3) Stam 24 190156 Y 1181200206 AT121590 1427 2-1-10-1-9-1-21 (4) Stam 29 200992 K 1181200207 AT121590 1427 2-1-10-1-9-1-21 (4) Stam 30 202323 O 1181200208 AT12189 15 1-1-1-1-1-1-1 (3)
Resistance genes K. pneumoniae Phenotype Found Resistance gene 18/18 aac(6')-ib 18/18 aada1 Aminoglycoside resistance 18/18 aadb 18/18 aph(3')-ic 18/18 stra 18/18 strb 18/18 blaoxa-9 Beta-lactam resistance 18/18 blashv-12 6/18 blatem-1a Fluoroquinolone and aminoglycoside resistance 18/18 aac(6')ib-cr Fosfomycin resistance 18/18 fosa Phenicol resistance 18/18 cata1 18/18 cmla1 Quinolone resistance 18/18 oqxa 18/18 oqxb Sulphonamide resistance 18/18 sul1 18/18 sul2 Tetracycline resistance 18/18 tet(d)
Comparison with existing methods
MRSA - comparison with established methods
Inter-laboratory reproducibility Twenty Staphylococcus aureus DNA samples DNA samples were prepared using the MagAttract HMW DNA kit (Qiagen, Hilden, Single sequencing run on an Illumina MiSeq Nextera XT library preparation kit and the 250-bp paired-end sequencing chemistry version 2
Run acceptance criteria Sequencing output 5.6 Gb (to achieve an average sequencing coverage of 100-fold for the 20 samples with genome sizes of 2.8 Mb) Q30 read quality score of 75% SeqSphere software version 2.4 or higher (Ridom GmbH, Münster, Germany) Default parameters for quality trimming, de novo assembly, and allele calling Specifically, reads were trimmed at their 5 -and 3 -ends until an average base quality of 30 was reached
Inter-laboratory reproducibility
Inter-laboratory reproducibility high reproducibility and accuracy of WGS-based microbial typing when using a standardized methodology
Wet-lab and e-lab only (Isolates, DNA, Fastq)
SNP
Conclusions GMI PT2014 The Wet lab part worked as anticipated and provided interesting data identifying several sequencing deviations such as contaminations and poor sequencing The PT indicated several quality markers which could be used and considered as future QC standards for assessing sequence quality With the limited data submitted in the pilot PT it was, however, not possible to determine specific QC measures
QC parameters @UMCG Number of contigs < 1000 N50 value > 15000 Maximum contig length > 50000 Coverage > 30x Percentage used reads > 90% % expected genome size > 90% en < 115%
QC beyond WGS 16S microbiome sequencing 16-23S diagnostic sequencing in clinical samples Metagenomics
16-23S Diagnostic sequencing DNA extraction direct from clinical material 16S-23S rrna PCR (~ 4.5 kb) Library preparation Next Generation Sequencing (MiSeq, V3, 600) de novo reads assembly (CLC Genomic Workbench, Qiagen) basic local alignment search (BLAST, le BiBi) Sabatet al, Sci Rep. 2017 Jun 13;7(1):3434. doi: 10.1038/s41598-017-03458-6.
Identification of pathogens in positive blood cultures Sabatet al, Sci Rep. 2017 Jun 13;7(1):3434. doi: 10.1038/s41598-017-03458-6.
Identification of pathogens in orthopedic samples Sabatet al, Sci Rep. 2017 Jun 13;7(1):3434. doi: 10.1038/s41598-017-03458-6.
Real-time PCR Setup Delftia DSMZ Clinical samples Run control Negative control Spiked with PhHV (DNA) or EMCV(RNA) (generic internal control) NA extraction ABI PRISM MISEQ Amplification Negative control Diagnostic Targets PhHV EMC Delftia DSMZ
Capacity building for interventional genomics in clinical microbiology NGS-16S/23S Metagenomics Transcriptomics NGS-Typing Sequence based typing PCR-based Typing
MetaNet Metagenomics for clinical microbiology Capacity building workshops (October 2018 Groningen ESCMID) Organize a ring test and proficiency testing (EQA, QC) Develop or improve databases for pathogens, host genes and know pathogen-host relations MetaNet
Acknowledgements Microbes in Health and Disease Post-docs and PhD students 16-23S Study Group / Remis + The professor Wet-lab Metanet CbAGS-net Capacity builders The Guests The Pioneer Genomics for IP