This document discusses how Public Health England is implementing whole genome sequencing (WGS) as the routine method for Salmonella reference microbiology. Salmonella causes many cases of foodborne illness in the UK each year. WGS provides higher resolution typing than previous methods and allows for global data sharing. PHE has invested in the infrastructure and bioinformatics capabilities needed for large-scale Salmonella WGS. WGS was used to investigate an international Salmonella Enteritidis outbreak. Analysis of outbreak strains identified multiple clusters corresponding to different outbreaks. WGS also determines antimicrobial resistance profiles and allows characterization of bacterial strains. PHE aims to continue improving Salmonella analysis methods using WGS.

1. Revolutionising Public Health
Reference Microbiology Using
Whole Genome Sequencing:
A Case Study with Salmonella
Philip Ashton
Bioinformatician, Gastrointestinal Bacteria Reference Unit

2. 2 Salmonella WGS for Public Health
PHE are using WGS as the
routine test for Salmonella
reference microbiology.
All Salmonella isolates are being
sequenced.
No routine serotyping.

3. Why?
How?
What?
Salmonella WGS for Public Health3

4. 4 Salmonella WGS for Public Health
Why do we want to
implement WGS
for Salmonella
reference micro?

5. 5 Salmonella WGS for Public Health
> 38000 cases of
Salmonella
gastroenteritis in
the UK per year, of
which > 11000
report to their GP.
Tam et al, 2011, Gut
gut.2011.238386
Typhoidal
Salmonella as well.
77 deaths in UK 2008,
FSA Foodborne
Disease Strategy

6. 6 Salmonella WGS for Public Health
Tom Innes & PT14b OCT
‘protecting and
improving the nation’s
health’.

7. 7 Salmonella WGS for Public Health
Previous typing method - serotyping
Satheesh Nair

8. 8 Salmonella WGS for Public Health
David Powell

9. 9 Salmonella WGS for Public Health
PHE Salmonella surveillance data - http://bit.ly/1CdEOe0

10. 10 Salmonella WGS for Public Health
PHE Salmonella surveillance data - http://bit.ly/1CdEOe0

11. Salmonella is a major public health issue.
Existing typing/subtyping has limited phylogenetic
insight.
Need higher resolution typing methods (WGS) for
monitoring trends and detecting outbreaks.
Unique opportunity for global data sharing
11
Summary – why WGS?
Salmonella WGS for Public Health

12. 12 Salmonella WGS for Public Health
How will WGS based
Salmonella reference
microbiology work?

13. WGS implementation at PHE
2011- 2015
2 MiSeq machines 2 HiSeq 2500 high-throughput machines
Capacity > 3,000 genomes per week
+
PHE investment in WGS: financial,
laboratory, bioinformatics, data
handling, staff training
=
Salmonella WGS for Public Health13

14. Identification / Mixed – kmer
gateway
K-mer Gateway
14
Acetobacter
Acinetobacter
Actinomyces
Aeromonas
Aggregatibacter
Bacillus
Bacteroides
Bartonella
Bifidobacterium
Bordetella
Borrelia
Brucella
Burkholderia
Campylobacter
Chlamydia
Chlamydophila
Clostridium
Corynebacterium
Desulfovibrio
Enterobacter
Enterococcus
Escherichia
Francisella
Fusobacterium
Gardnerella
Gordonia
Haemophilus
Helicobacter
Klebsiella
Lactobacillus
Legionella
Leptospira
Leuconostoc
Listeria
Morganella
Mycobacterium
Mycoplasma
Neisseria
Nocardia
Paenibacillus
Prevotella
Propionibacterium
Pseudomonas
Rhizobium
Rhodococcus
Rickettsia
Salmonella
Shewanella
Shigella
Staphylococcus
Streptococcus
Streptomyces
Treponema
Ureaplasma
Vibrio
Yersinia
Off all the k-mers of length 18
in each reference genome
what percentage are in our
sequencing reads?
Can be used to identify cross
species contamination.
Salmonella WGS for Public Health
99.7% accurate
subspeciation

15. 15 Salmonella WGS for Public Health
Serotype inferred
via MLST
Satheesh Nair
Achtman et al., 2012

16. MLST to infer serotype
16
Method
Short Read
Sequence Typing
ST (and eburst
group)
Serotype
Result
Inouye et al., 2012
6887 strains with
WGS and pheno
6616 (96%) results
matched
Salmonella WGS for Public Health
2 serotypes – 1 ST/EBG
Lab error
No ST/serotype lookup

17. Novel Sequence Types
17 Salmonella WGS for Public Health
SISTR/SeqSero

18. 18 Salmonella WGS for Public Health
Challenges:
• Many different eburst groups (of STs) – have to be analysed separately
• Hundreds of strains a week
• Rapid, hands-off analysis
Solution – SnapperDB:
Most common serotypes – SNP typing
Sample
FASTQs
(with ST)
EBG 1 - Typhimuriumdb
db
db
db
db
EBG 4 - Enteritidis
EBG 13 - Typhi
EBG 3 - Newport
EBG 11 – Paratyphi A
…
30 mins - parallel 5 min – 1 hour

19. 19 Salmonella WGS for Public Health
What will be the output
of WGS for Salmonella
reference micro?

20. • Example of use in outbreak – PT14b
• Strain characterisation – antibiotic
resistance profile
• Future perspectives
20 Salmonella WGS for Public Health

21. Salmonella NGS at PHE
Salmonella Enteritidis Phage Type 14b
International outbreak
Tom Inns/PT14b Outbreak Control Team – Inns et al., Eurosurveillance, 2015

22. 22 Salmonella WGS for Public Health
0.02
*
*
*
OUTBREAK 4
OUTBREAK 2
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 2
OUTBREAK 3
OUTBREAK 5
*
OUTBREAK 2
OUTBREAK 3
OUTBREAK 3
OUTBREAK 1
OUTBREAK 4
*
*
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
OUTBREAK 1
*
OUTBREAK 4
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
*
*
OUTBREAK 1
OUTBREAK 2
OUTBREAK 1
OUTBREAK 4
*
OUTBREAK 1
OUTBREAK 3
OUTBREAK 4
OUTBREAK 2
OUTBREAK 4
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 1
OUTBREAK 3
OUTBREAK 3
OUTBREAK 5
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 1
OUTBREAK 4
OUTBREAK 3
OUTBREAK 2
*
OUTBREAK 2
OUTBREAK 3
OUTBREAK 2
OUTBREAK 4
*
OUTBREAK 3
*
*
OUTBREAK 3
*
OUTBREAK 3
OUTBREAK 3
*
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
OUTBREAK 4
OUTBREAK 3
OUTBREAK 3
OUTBREAK 2
OUTBREAK 2
OUTBREAK 3
*
OUTBREAK 2
OUTBREAK 5
*
OUTBREAK 2
*
*
OUTBREAK 1
OUTBREAK 1
OUTBREAK 3
OUTBREAK 2
*
OUTBREAK 2
OUTBREAK 1
OUTBREAK 3
*
OUTBREAK 2
*
OUTBREAK 2
OUTBREAK 1
*
OUTBREAK 3
*
OUTBREAK 1
*
*
OUTBREAK 1
OUTBREAK 3
OUTBREAK 3
*
*
OUTBREAK 3
OUTBREAK 3
*
OUTBREAK 1
OUTBREAK 3
*
OUTBREAK 4
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 1
*
OUTBREAK 4
*
OUTBREAK 3
*
OUTBREAK 3
*
OUTBREAK 3
*
*
*
OUTBREAK 1
OUTBREAK 2
OUTBREAK 3
OUTBREAK 3
OUTBREAK 2
*
OUTBREAK 4
OUTBREAK 1
*
OUTBREAK 1
*
*
*
OUTBREAK 3
*
*
*
OUTBREAK 2
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
*
*
OUTBREAK 4
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
OUTBREAK 1
OUTBREAK 3
OUTBREAK 1
OUTBREAK 4
*
OUTBREAK 5
OUTBREAK 1
OUTBREAK 3
*
OUTBREAK 4
*
*
*
*
OUTBREAK 1
OUTBREAK 3
OUTBREAK 3
OUTBREAK 4
OUTBREAK 2
OUTBREAK 2
*
*
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
*
OUTBREAK 3
*
OUTBREAK 4
*
*
OUTBREAK 5
OUTBREAK 4
*
*
OUTBREAK 2
*
OUTBREAK 3
*
OUTBREAK 2
*
OUTBREAK 2
OUTBREAK 4
OUTBREAK 4
OUTBREAK 5
*
OUTBREAK 5
OUTBREAK 3
*
OUTBREAK 4
OUTBREAK 1
*
OUTBREAK 4
OUTBREAK 4
OUTBREAK 1
*
*
*
OUTBREAK 5
OUTBREAK 2
*
OUTBREAK 1
*
*
OUTBREAK 3
OUTBREAK 1
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
*
OUTBREAK 3
*
*
OUTBREAK 1
OUTBREAK 1
*
*
OUTBREAK 1
*
*
OUTBREAK 1
OUTBREAK 4
*
*
OUTBREAK 3
OUTBREAK 4
*
*
*
*
OUTBREAK 2
*
*
OUTBREAK 4
OUTBREAK 3
*
*
OUTBREAK 1
OUTBREAK 1
*
*
OUTBREAK 4
OUTBREAK 3
*
OUTBREAK 4
OUTBREAK 3
*
OUTBREAK 3
OUTBREAK 4
OUTBREAK 2
OUTBREAK 1
*
*
OUTBREAK 4
*
*
OUTBREAK 4
OUTBREAK 4
OUTBREAK 4
OUTBREAK 1
OUTBREAK 3
*
OUTBREAK 1
OUTBREAK 1
*
*
*
OUTBREAK 2
*
*
OUTBREAK 3
OUTBREAK 2
*
*
OUTBREAK 3
OUTBREAK 1
OUTBREAK 4
*
*
OUTBREAK 2
OUTBREAK 2
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
OUTBREAK 2
*
*
*
OUTBREAK 1
OUTBREAK 3
*
OUTBREAK 3
*
OUTBREAK 4
OUTBREAK 5
OUTBREAK 1
OUTBREAK 1
*
*
*
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 1
*
*
OUTBREAK 3
*
OUTBREAK 4
OUTBREAK 1
OUTBREAK 2
OUTBREAK 3
*
OUTBREAK 3
OUTBREAK 1
OUTBREAK 3
*
*
*
OUTBREAK 4
OUTBREAK 1
OUTBREAK 1
OUTBREAK 1
OUTBREAK 3
*
OUTBREAK 4
OUTBREAK 4
OUTBREAK 3
OUTBREAK 1
*
*
*
OUTBREAK 3
*
*
*
OUTBREAK 2
*
OUTBREAK 4
OUTBREAK 1
81094_H14372077201-1
OUTBREAK 2
OUTBREAK 2
OUTBREAK 3
*
*
OUTBREAK 1
OUTBREAK 2
*
OUTBREAK 3
OUTBREAK 3
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 2
OUTBREAK 1
*
OUTBREAK 2
OUTBREAK 3
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
OUTBREAK 1
OUTBREAK 4
OUTBREAK 3
*
OUTBREAK 1
OUTBREAK 3
*
*
*
*
OUTBREAK 4
OUTBREAK 3
*
OUTBREAK 1
*
*
OUTBREAK 3
OUTBREAK 4
*
*
OUTBREAK 3
*
*
OUTBREAK 1
*
*
OUTBREAK 1
OUTBREAK 2
OUTBREAK 4
OUTBREAK 4
OUTBREAK 2
OUTBREAK 3
OUTBREAK 1
*
OUTBREAK 3
OUTBREAK 3
OUTBREAK 3
OUTBREAK 2
*
OUTBREAK 1
OUTBREAK 3
OUTBREAK 3
OUTBREAK 2
0.02
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Hospital – W.Mids
Kebab Grill - London
Chinese Restaurant - Wessex
Chinese Restaurant - Cheshire
Chinese Restaurant - Lancashire
Tim Dallman, Tom Inns, Thibaut Jombart

23. Hierarchical clustering and
Geography
23
Alison Waldram & Gayle Dolan

24. 24
Phenotype S Phenotype R Error%
Genotype S Genotype R Genotype R Genotype S
CHL 580 2 39 3 0.78
SUL 466 1 149 8 1.40
TET 467 3 151 3 0.93
TMP 562 3 57 2 0.78
AMP 484 5 134 1 0.93
CTX/CAZ 618 1 5 0 0.16
CPR 619 1 4 0 0.16
FOX 612 0 7 5 0.78
CIP 478 4 138 4 1.25
NAL 485 2 127 10 1.87
GEN 608 1 14 1 0.31
TOB 613 2 8 1 0.47
AMK 622 0 1 1 0.16
STR 486 18 119 1 2.96
7700 43 953 40 0.95
AMS: Genotype vs phenotype
Major errors Very major errors
Michel Doumith/Martin Day
Validation : 642 Salmonella strains
Resistance : 57.5 % susceptible
24.7 % multi-resistant (> 2 classes)
Salmonella WGS for Public Health

25. Future perspectives
25 Salmonella WGS for Public Health
• Move from MLST to kmer/phylogenetic placement for
sorting before SNP analysis, especially for Enteritidis.
• MinION sequencing:
• Outbreak investigation (Quick, Ashton et al., Genome
Biology, 2015, v16, p114),
• Solving complex + interesting regions of the genome
(Ashton, Nair et al., Nature Biotech, 2015, v33, p296)
• Live streaming of sequencing data – google ‘Loman
labs’
• National (One Health) and international (GenomeTrakr)
real-time collaborations

26. Summary
• WGS is a single test – multiple assays
• Better quality information than existing
tests
• Evolutionary and phylogentic insight
Salmonella WGS for Public Health26

27. Acknowledgements
27 Salmonella WGS for Public Health
Microbiology
Elizabeth de Pinna, Tansy Peters, Satheesh Nair, Martin Day,
Anna Lewis, Tim Dallman, Kathie Grant and other staff in the
Salmonella lab
Epidemiology
Alison Waldram (FETP), Richard Elson, Chris Lane, Tom Inns
Genomic Services Unit
Cath Arnold and team
Bioinformatics Unit
Jonathon Green, Anthony Underwood, Rediat Tewolde, et al.