In April 2016, SfAM held its annual spring meeting on the topic of whole genome sequencing (WGS), asking “What can it do for clinical and public health microbiology?”
With experts from many fields both giving presentations and in the audience, they took the opportunity to make a change to their usual proceedings and finished the day with a lively panel discussion. While you really had to be there to get the full experience, we’ve presented below a summary, alongside some of our favourite quotes.
Our panel consisted of our speakers for the day:
- Dr Anthony Underwood from Public Health England
- Professor Noel McCarthy from the University of Warwick
- Professor Edward Feil from the University of Bath
- Dr Grace Smith of the Heart of England NHS Foundation Trust
- Professor Judith Breuer from University College London
- Dr Alison Mather from the University of Cambrige
- Professor Mark Fielder from Kingston University (The Chair for the panel)
Audience: Are there barriers or pitfalls to applying WGS to the clinic?
Grace Smith: It’ll take a long time – not physically difficult, the difficulties are interpretation
Judith Breuer: Especially for viruses as we can’t look at the phenotype
Edward Feil: Predictions are spectacular for bacteria – “I don’t see [WGS] as a replacement [for phenotypic testing]”
Alison Mather: There is value in discovering new resistance data
Judith Breuer: “Information at the bedside”
Audience: “What does this mean for strain collections?”
Judith Breuer: Its already happened with virus – they’re gone. It needs to be cost effective and if nobody is using the collection then it won’t be maintained.
Audience: What is the sensitivity of WGS directly from sample?
Judith Breuer: “Will be as sensitive” as PCR identification for Noravirus – PCR has mismatches due to variability of viral genome, it may not replace PCR due to the cost of WGS vs PCR but as technology develops it may become a rival.
Anthony Underwood: “How well will it apply to bacteria?”
Judith Breuer: There are issues with mobile elements in gram negatives but nanopore addresses these issues. It’s not there yet but we can do variable RNA viruses so there’s scope.
Audience: How to deal with huge numbers of bacteria?
Judith Breuer: “Different approaches to different technologies”. Metagenomics and deep sequencing give us lots of info. It may not be important to get every sample of bacteria in a population.
Same audience member: But we have found many strains of antibiotic resistance that aren’t present in the majority but in the minority of a population
Grace Smith: We’ve missed the majority for years, in reality we’re actually getting closer to identifying the minority in a sample, not the other way around.
Mark Fielder: And as tech improves we’ll get closer to identifying those anomalies.
Anthony Underwood: Now we can identify some of those minorities we missed before and investigate them
Audience: What do you think of WGS over MALDI-TOF for diagnostics?
Anthony Underwood: We now have slightly more resolution over MALDI
Grace Smith: MALDI-TOF is faster than WGS so we need to address that to be diagnostically relevant
Andrew Underwood: WGS is not cheap for TB
Grace Smith: Many larger labs do it cheaper themselves
Judith Breuer: It’s likely WGS and MALDI-TOF will overlap – use the quick MALDI TOF to predict what needs WGS
Audience: Is there any special procedure for GC rich organisms?
Anthony Underwood: There are kits optimised for about 50:50 AT:GC but more GC leads to worse coverage
Audience: On MALDI TOF, it isn’t perfect for every species – wouldn’t trust it with Strep. WGS fills in the shortcomings of MADI-TOF, which is “limited to the 16S window”.
Audience: WRT resistance in the food chain, WGS can’t predict level of resistance, just presence or absence of resistance genes
Judith Breuer: In virology we know which mutation give low or high level resistance and can identify these.
Grace Smith: We also have this in TB. It’s early days yet.
Alison Mather: It’s easier with SNP resistance but acquired resistance genes make it not so easy.
Noel McCarthy: We need a long time to get enough data on patients to back up what WGS might be predicting.
Edward Feil: Generally we are slowly getting there. Phenotypic testing isn’t perfect anyway.
Alison Mather: Also copy number – a high copy number can turn genes into resistance genes where they weren’t before, and WGS doesn’t show this.
Anthony Underwood: WGS can help eliminate human errors in phenotypic testing, which makes WGS an attractive concept for antimicrobial resistance identification
Audience: Do we need to redefine Koch’s Postulates?
Judith: Koch’s Postulates refers to disease so should still be valid. All new viruses identified by WGS fulfil cox postulates.
Audience: Concerned not addressing info to clinicians themselves, so how can we get the ‘blah blah’ down to definitive answers?
Grace Smith: We need a proper discussion about the patient themselves in difficult cases, therefore we have to engage the clinicians.
Judith Breuer: It depends on the patient group, and the tech is still in development and finding out where it can all fit in. It’s not always clear how WGS info can be useful at first glance
Anthony Underwood: We shouldn’t be scared of expressing things as a probability, therefore we need to draw mathematicians into the field for this interpretation.
Noel McCarthy: Let’s not underestimate the clinicians’ ability to adapt. They engage with useful technology so if WGS is useful, they’ll engage with it.
Audience: Will we ever say goodbye to petri dishes and microscopes?
Judith Breuer: It’s already happened in Virology
Audience: And will WGS ever be cheaper and quicker than conventional techniques?
Judith Breuer: It’s possible we could get WGS in 10 minutes – “Would you wait overnight for a plate?”. Look how quick PCR has become. And it might not even be WGS that replaces these techniques.
Mark Fielder: And as technology improves, it gets funded better.
Audience: How should I train the microbiologist of the future?
Anthony Underwood: With a programming course
Grace Smith: Add WGS into the course, but don’t use it to replace anything
Judith Breuer: Don’t be afraid to discuss where genomics fits into other settings
Mark Fielder Sums up: “We’re in an exciting time. We’re not there yet but good things are coming… It may not be tomorrow but it’s like to come.” Need useful tools in the clinic so we’ll get better diagnostics and therefore we’ll make more useful tools.
Categories: SfAM News