Cracking superbug codes: can a global genetic map save us?

One of the people trying to find a way past AMR is Andrew Singer, principal scientist at the UK Centre for Ecology & Hydrology. He takes the view that, to tackle a global problem, which this very much is, you need a global solution. And it turns out that, all around the world, a great number of disease outbreaks caused by antibiotic resistant microbes are being genetically sequenced to find out what the organisms are, and where they came from. But these data are only very rarely shared, meaning we struggle to see the big - joined up - picture of how antimicrobial resistance is operating and shifting on worldwide scales. And, as Andrew argues, prevention is much better than cure, and if we better understand the source of things like drug resistant E.coli infections, we’re better placed to stop it…

Andrew - What we can do is we can have all of these people across the world submit sequences of E. coli to a database and that database will accumulate all of these sequences. And as their sequence is being submitted to this database, it asks the database, where did this come from? Now, you know where your sample came from because it came from a human or came from the environment, but you don't know where that E. coli came from. And that's the critical bit, because if you know where it came from, you could do something about it. So was the bathing water polluted with sewage or was it polluted by a dog that pooed before you showed up or was it contaminated by a bird that just happened to really time itself really well? All of that matters because if you're the environment agency in this case, you have to solve that problem. And you can't solve it if you don't know where it came from. And if it's a natural thing, then you might say to people, don't feed the birds or don't walk your dogs on the beach. If it comes from sewage, you have another problem that needs solving. Or if it comes from farms and you know that it rained a lot and that the water rushed off the farm and it came into the river and the river contaminated your beach, you can then pinpoint, OK, there's something that needs to be done here.

Chris - So the genetic codes of these microbes is sufficiently different across those different situations, a source from an animal, a source from Thailand versus London. They're sufficiently different that we will be able to tell the difference if we've got those sequences. And we can almost plot a map of the world and the different example cases of what microbe goes with what. And that will give us that giant dictionary almost of what infections probably originate from what sources.

Andrew - Exactly. So at the moment, we don't have this, what we call a trusted research environment, which is what we're building. And it's a basically a playground for data. And the data sets that we currently have are really quite limited in numbers. They're in the thousands. What we need are millions of sequences, all E. coli from all over the world, from all of these different sources, from birds to dogs, to humans, to cows, to pigs, from everything, from every human that we can get a sample from. And you'd be surprised how many sequences are being generated a year just in the UK. It's hundreds of thousands. And so if we have a database where all of this goes and it's annotated who it came from, and then we train the algorithms to understand what a sequence looks like when it comes from a human, the likelihood of it then finding a new sequence or being asked, why did this new sequence come from? It's a very high likelihood that it's going to pinpoint where it came from and where it's not sure where it came from. If it came from a farm and it came from a farm worker, you are then asking the question, well, this is a really risky farm because it has E. coli that is able to transfer from a cow to a human or a human to a cow. And you don't really want that happening because that's how pandemics start. So you want to nip that one in the bud.

If it's something that's always human or always cow, then you feel a bit safer and you think, well, someone just did something a bit silly.

Chris - How does this solve the problem of antimicrobial resistance? If we've got that map and we can say, very well, I know where this thing came from. That doesn't help the fact that I'm still dying of infection I can't treat.

Andrew - Essentially, if you can problem solve, so if you know where the infections are coming from, the best way to combat AMR is to prevent people from getting infected in the first place.

Chris - So this is prevention.

Andrew - Yeah. So I've been led to believe that prevention is not the sexiest thing on the block. However, it is probably where we should be spending our time. This is a global resource. So once we all join this, we all benefit by having more and more data.

Chris - Presumably, then, if we know where something came from and it also happens to be an antimicrobial resistant form, then that gives us enormous information about sources of infection and spread of infections that are more likely to be troublesome.

Andrew - Yes, it's a canary in the coal mine kind of approach of trying to do prevention where you find the infectious organism as quickly as you possibly can. And once you've found it, you then interrogate its genome, looking for the resistance traits. And then depending on where in the world that had come from, you then have an added bit of information to then tackle the problem of AMR in the sense that if it came from another country, then you have maybe a higher priority of where investment needs to go.

Chris - But how far down the path have you got so far? And have you got buy-in from the international community? Is everyone willing to sign up and do this? Because I mean, it sounds very laudable, but it's going to be expensive, because this is not cheap doing this. And you need good technology to do it, which might limit the places on earth that will buy into the project. And it's going to take time. When do you think that you'll be able to walk into a room like this one where you and me are chatting, and show me this effective giant radar screen for the spread and origin of bacterial infection, plus or minus whether or not they're resistant to infections from the entire world?

Andrew - The proof of principle will be probably demonstrated in a number of use cases in the next couple of years. At any point during that next couple of year journey, the funding then needs to show up to then support this trusted research environment, which would then be a global resource. I'm always very optimistic, which puts me in around three to five years. It's too obvious of a solution for us to pass up. It's so good. And we're also in the driver's seat, we have all of the resources we need to do this in this country, and then be the leader. And I think we should just take it because I don't think it's that expensive.