Hello, and thank you for listening to the MicroBinFeed podcast. Here we will be
discussing topics in microbial bioinformatics. We hope that we can give you some
insights, tips, and tricks along the way. There's so much information we all
know from working in the field, but nobody writes it down. There is no manual,
and it's assumed you'll pick it up. We hope to fill in a few of these gaps. My
co-hosts are Dr. Nabil Ali Khan and Dr. Andrew Page. I am Dr. Lee Katz. Both
Andrew and Nabil work in the Quadram Institute in Norwich, UK, where they work
on microbes in food and the impact on human health. I work at Centers for
Disease Control and Prevention and am an adjunct member at the University of
Georgia in the U.S. Hello and welcome to the MicroBinFeed podcast. Andrew and I
are your co-hosts today, and we're continuing our series where we examine a
particular microbial species in some depth. Today, again, we're going back to
mycobacterium tuberculosis, and we're hoping we can discuss some of the specific
issues for bioinformaticians to keep in mind when studying this organism.
Joining us again, our guests are Dr. Susie Hingley-Wilson, who is a lecturer in
bacteriology at the University of Surrey. She works on tuberculosis, and her lab
focuses on survival and persistence. We also have Dr. Jani Best, who is a senior
lecturer in microbial metabolism at the University of Surrey. She is a bacterial
dietician, so she's interested in what MTB actually eats. And we are also joined
by Dr. Connor Meehan, who is an assistant professor in molecular microbiology at
the University of Bradford. He specializes in whole genome sequencing and
molecular epidemiology of pathogens, particularly mycobacterium tuberculosis,
and also does genome-based bacterial taxonomy. So let's start off with Andrew,
who wants to tell us about software we've been developing. So I have actually
developed software for TB, bioinformatics software for TB, which is quite a
surprise to everyone I know, and quite shocking. We have a tool called GalRoo,
and what it does is polygotyping on long reads. Originally, though, it was meant
to do something totally, totally different, but we reused the code when we
figured out that wouldn't work, and it did work for TB. So thank God, and we
have a preprint out of that. So it just shows that you should never throw away a
failed project. You can always recycle it in something else. But what it does is
it just looks at long reads, does a bit of polygotyping. It produces some kind
of numbers that people who like TB seem to be interested in. I don't know much
about it, to be quite frank. But I understand there are many other
bioinformatics tools for TB, and so maybe Conor can give us an idea of what's
going on there, because I know you've written a review recently. Yes, well, 40
people wrote that review. I just coordinated them and did some of the writing
myself. But bioinformatics in TB is very much a field that is changing by the
day. TB was very clinician driven for a very, very long time. So what a lot of
people think of as being basic biology and basic bioinformatics is kind of
coming in at the end and then trying to fit into the systems that we saw before.
So GalRoo, which is a good tool, and I've used it, and it's good, I like it. It
does polygotyping on long reads. For people that don't know polygotyping, it's
essentially CRISPRs. It's a presence and absence of spacers in the genome. It's
called polygotyping in TB because it is, but it's essentially CRISPRs. And it's
the presence or absence of 23 of these that will tell you whether it's a certain
lineage or something else that's out there. So it's a lot of the tools now are,
well, in the clinic, we did it this way. Now can we get that information from
the genome? And then a lot of them were built for the short read one. And now
it's moving into the, can we do it with the long reads? But obviously
bioinformatics, to a lot of people who are using the tools, they're like, but
this is the same, but obviously bioinformatics, we know that this is a
completely different program in order to do that. So an example I'll give you
is, there's a lot of tools like TB profiler, micro-TB, are the ones that will do
drug resistance from it, but they all start with short reads. So now we have to
remake all the tools, so they'll work with nanopore reads or something. So TB
profiler will now work on the nanopore reads. But a lot of the bioinformatics
tools are really, really based around the short reads. We use specific SNPs to
tell you what lineage it is. We have specific SNPs to tell you what drug
resistance it is and all of that. And we've got to port all that now to the long
reads. So it's been a field that's massively exploding outwards. And now a few
of us are trying to work to kind of bring it back in line and say, well, here
are the things we want from a tool that does this. And the review was really
trying to say, what are the tasks that we need genomic data to do? And then how
would we know that a tool is ticked off to say that it does that thing? What is
the database of SNPs you're using? What is the technology you're using? How are
you calling a minority variant? Because a minority variant is very different to
different people. Some tools say 90% minimum cutoff of the reads at that
position to call it the majority. Others say 70. Others then are saying we only
need two reads for a minority variant. Other ones are saying five reads. So
we're trying to kind of say, well, here are some standards, you know, the rules
help control the fun. And that's where we're going with the tools at the moment.
And so what kind of questions would people like Danny and Susie like to ask of
genomics? If you could have anything you wanted to back up your wet lab
experiments? Well, I guess one thing I'm quite interested in, and I kind of
tried to get a grant funded on this, but the MRC weren't as interested as me in
it, was that question I kind of talked about before, which is, you know, what
are the drivers of the evolution of drug resistance really in TB? You know, what
sort of environmental conditions may drive that to a particular evolutionary
trajectory? I'm kind of really, they've done some really cool stuff on E. coli,
where you look at different sort of metabolic environments and you see that
there's a different evolutionary trajectory. And I'm kind of interested, because
obviously, when you're looking at human, it's really complicated and you don't
know what the different conditions are. So I'm kind of I'm kind of interested in
that sort of basic biology question, which may have significance, because
obviously, if you could sort of break that evolution or, you know, in some
manner, this might be an approach to making the drugs last longer with this kind
of adjuvant therapy. So I'm kind of quite interested in that approach. And
something else that's very interesting, I know Susie's very interested as well,
is these mutations which actually allow strains of TB to survive for longer in
the presence of drugs. And they might not make them resistant. They might just
make them more tolerant. And then they have the ability to go on to become
resistant. So those are kind of two things that I kind of wanted to kind of look
at and I'm quite interested in. So has anyone ever looked at which genes are
essential for the life of MTB, like using Trades or anything like that? Yes,
there have been others. There's some studies by a group in Aberystwyth
University who've looked at, they've used something called Trades, which uses a
set of mutants and then they look at survival. They've looked at mycobacterium
tuberculosis and also mycobacterium bovis, which causes TB in cattle and can
also cause TB in humans. So yeah, that has been, that hasn't been published yet.
Hopefully, you know, that should come through soon. But that would be really
interesting. But there's multiple transposal library screens done in TB and
macrophages and they're starting to do a little bit of macaque models in
different, all sorts of different conditions. So there's a lot of information
and a lot of arguments actually in the field about, you know, what then
constitutes a good drug target? Does it have to be an essential gene? And that
to me is always an essential, it's always conditionally essential because I
always find that weird, right? So it's essential for life in a lab on the, on
it's, you know, it's not really necessarily going to be what's essential in the
host. So there has been a lot of those studies in a variety of different
conditions. So Danny, anything else that you'd like to add from your wish list?
Well, I guess answering that, that question about the evolution, the mutation
rate of TB would be kind of interesting because I don't think we actually, you
know, there was a study done where we really haven't, is that connected to
growth rate or not? That would be kind of something with fundamentals that I
think would also hasn't been answered in TB. What is, you know, what is that
mutation rate and how does it differ inside the human body? So we have a lot of
different lineages in TB. They don't have fancy variant names like we see in
COVID. It's just one to nine. Mycobacterium tuberculosis is a species and we
refer to it as mycobacterium tuberculosis complex. The diversity that we know of
is growing and growing and growing. So up until a few years ago, we only knew of
six lineages. Then a seventh one came out in Ethiopia. We found one, which we
call lineage eight, then in Rwanda and Uganda. I was also part of.  the one that
found lineage nine. So it's growing and growing and growing as you go along. And
some studies have looked at the mutation rates of these different lineages and
found that they are slightly different between some of the lineages. Tend to
hover around the 10 to the minus six, 10 to the minus seven, which is per
generation, right? So per replication cycle, but in TB, that is a very slow one.
So we tend to average out at about one snip every three years if it's left by
itself. So this makes it very difficult when you do epidemiology because you'll
just end up with zero snips between two different strains. And then you have no
idea if they transmitted to each other. What I will say is it seems to, from
this is using the genome to figure out the mutation rates. And the difficulty
there is that it's more accurate to say that we do near whole genome sequencing
because we throw out about 10% of the genome at the moment in repetitive regions
so that we have PPE and ESX genes and stuff like that. So we're throwing those
out and there may be more mutations that are occurring in there. So the mutation
rate isn't even across the whole genome. It might be that there's more mutations
in some sections and less in others. So the genomics is helping us understand
that, but because the mutation rate is so slow, we need data sets that go over
decades in order to be able to tell that. And those are very, very rare. They're
very, very difficult to capture those big data sets. So it's coming, but we just
need large sequencing over longer periods of time to try to look at that for the
mutation rate at least. Can I ask a question then? So what do you say the
generation time is? Because that must be a massive fudge, right? So this 10 to
the minus six, 10 to the seven mutation rate is based on a generation time of
what? A day, isn't it, I think, or specifically? But that's in the lab, right?
Yeah, yeah, exactly. And I think from some mice studies, it's also from that.
But this is the difficulty that we have with micro-rectume tuberculosis, is that
we don't know as much of the fundamental biology as I think people assume we do.
When I came into the field in 2014, I was just like, oh, so we know all these
things. And they were like, no, we don't know any of those. I'm like, but this
has been around for so long. Having come from the HIV field, I was like, how do
we not already know these things? Like how to grow it in the lab? So we work
primarily on meninges five and six, which are only found in West Africa. They
grow micro-aerophilically and not aerophilically like the other tuberculosis
ones. And we found that this is due to a lot of different things in their
genome. So to get back to the first question you had, we can look at the genome
in order to say, well, in central metabolism, it seems to be a little bit
different. But testing that out is very, very difficult because you can test it
in the lab, but lineages five and six grow even slower in the lab. And it
doesn't always represent what we see in the patient, for sure. Well, in the
patient, a lot of growth is in the macrophage as well, which Danny's done some
really cool stuff looking at what TB eats within the macrophage. And then you've
got the level of heterogeneity there as well like what macrophage are you
looking at? So within the lungs, you've got alveolar macrophages. You'll have
naive cells coming in. So you'll have this huge heterogeneity in the macrophage
population. Then you've got TB grown within that. And then there's extracellular
TB. So there's this huge sort of variation as well, just to throw another
stunner in the works. Yeah, I'm curious what if, I mean, a macrophage isn't a
very pleasant place to be. What does it eat? It's worse than that. It lives in a
phagosome. So it's living not just within one membrane, but then it goes and
lives in the phagosome. So the very thing that's made to actually kill it. And
the kind of, I mean, there is a reasonable amount of different nutrients in that
phagosome that it can eat. But the kind of dogma is that it's mostly eating
sterols and fatty acids. Basically what the macrophage does is in response to
TB, it produces loads of sterols and that affects the immune system. So there's
a bit of a thing that we always puzzle about is it eating these things as a sort
of protective thing or does it really have to eat them? Certainly if you stop it
being able to eat fatty acids, it doesn't survive very well in a macrophage.
Yeah, it does have access to a number of amino acids in addition to that. Yeah,
I mean, TB is very fatty because it's got this bonkers cell wall. I don't know,
there's a famous paper, what is it called? That lazy bacteria. Yes, it eats and
it copes itself in, a lot of its genome is dedicated to genes involving fatty
acids, biosynthesis and fatty acid metabolism. But yeah. That's more or less
been my strategy for coping with lockdown. We do. Genomically, we've tried to
look at this from comparative genomics and we look between the lineages and we
see differences like in B12 and the use of pyruvate and all of this is
important. But a difficulty with tuberculosis is we don't have a lot of
neighboring species to compare it to that we have a lot of information. So we
have the mycobacterium genus, but most of the research has been done on
mycobacterium tuberculosis. And as we can see from the discussion, even that's
not giving the fundamentals. So we don't have a lot of nearby other pathogens
that we can do comparisons to because most of the nearby pathogens are the
environmental opportunistic pathogens of the mycobacteria. And they've all just
been based on the assumptions of what we know in TB. And then you have to go
pretty far to find another pathogen that maybe we know a lot about. So it's a
lot of fundamental stuff that needs to occur to generate that data by itself.
And by interrogating more whole genomes, we're able to get to that stage. And
there's some really cool things, like for example, BCG and TB, we found
metabolic differences that you can't actually explain by the genome, probably
not what we wanna hear on this podcast, but that actually BCG has the same
genes, but there's obviously something that's also happening, which means that
BCG, despite having the right genes to grow in whatever it is, is behaving
differently. So work is moving to the transcriptome level and the proteome level
now. So we're getting there with the genome, we're starting to move towards long
reads and finding gene differences and stuff. And now some people are getting
that money, thankfully, to start looking at transcriptomes. But we tried to do a
transcriptome, work for something else on TB, and we're like, oh, we'll just
gather all the public data sets of it. And it was just crickets. There just was
no public transcriptome data really to work with. Now we need more people to
generate them so we can start at least generating ideas of where we think things
would be different. Now I wanna just touch on that. The mycobacterium genus is a
pretty mixed bag of all sorts in there. I mean, you've got leprosy and you've
got embovis that we've talked about. You've got ulcerans, you've got, what's
that weird one, smeg, smeg, smegmatis? Yeah, it's just used a lot in the lab.
But in terms of the mycobacterium tuberculosis complex, there's like a bunch in
there. They're not something comparable. You can't, we can't borrow. I'm sure
that it would be nice to work on some of these other organisms to at least get
out of the category three restriction. So leprosy doesn't, you can't culture
leprosy. No. So you need a lot of rabbits if you wanna work with leprosy. And it
has a highly reduced genome because it lives intracellularly. Or squirrels, I
think. I think maybe. I think red squirrels. Yeah, that's leprosy, yeah.
Smegmatis is used a lot in the lab, but an example is this B12. A paper came out
recently that showed, yes, smegmatis can do it. It can create its own B12, but
tuberculosis cannot create its own. And they're still not even sure why, because
a lot of the pathway is there in the genome, but it just doesn't do it at the
end. It seems to import it. And then the questions we have is, okay, who is it
importing from? Because there's very few other bacteria that are living in these
phagosomes with them. So who's creating the B12 and the pyruvate and everything
for the TB to- We made a genome scale metabolic model of TB. And recently I had
a fantastically talented scientist who came from Columbia, and he basically
curated all the genome scale. Because we made the first one at Surrey, and
subsequently loads of people have made them. And B12 was really interesting,
because if you give the in silico genome scale model B12, it behaves more like
TB does. So in the end, we decided to let it make B12. And one of the reviewers
comments was exactly what you said, we don't think. Of course, that paper was
really clever, because all they did was just measure B12, didn't they? Why
nobody ever did that before? But I guess that doesn't answer the question if
there's in some special circumstances, TB can make B12, because it has the whole
set of genes to make B12. It is completely baffling. So again, it's like a
really fundamental, because B12 is really important for multiple pathways. But
the comparative genomics to the other opportunistic or obligate pathogens in the
same genus becomes a circular thing, because a lot of their genome based things
are based on microactivity tuberculosis in the first place. Or there's just not
a lot of people working on it. I can probably list you all the people who work
on ulcerans like from a...  lab point of view already because I've already
worked with them. So it's just not a lot of money in those as well either, so
they can be difficult to do. Which would be the analogue then outside of
mycobacterium that that you would suggest is the closest in terms of its
behaviour? Behaviour in what way I guess? So one organism that was used to
discover something useful was Rhodococcus. So Rhodococcus was used and it was
the pre-work to discover that TB was able to metabolise cholesterol, which is
quite a mad thing for a bacteria to be able to do. You know this great big
carbon molecule and that pre-work was all done on Rhodococcus. So you know there
is a group that do parallels, Rhodococcus is much more tractable in the lab. But
other than that I guess sometimes I look in Carinibacteria on the genetic, I
don't know if it's got you know for annotations for genes which are unannotated
to see if there's homologues in Carinibacteria. We've started to use some
containment level 2 mutants as well, which are auxotrophic mycobacterium
tuberculosis mutants that are made by Bill Jacobs lab in New York. So we've
started to use them because we look at a lot of microscopy, like live cell
imaging of single cells. So that's impossible to do unless you've got really
expensive equipment in the country, which we don't have. So we have started to
use some containment level 2 just on a separate side. So your closest genera are
Nocardia, Rhodococcus, Carinibacteria, Gordonia. They're the other ones in the
same family. It's also pretty sparse around that area of the whole tree. But
it's getting better as more money is going in and we're able to do the genomes
better and then at least pose questions to then go back in. We just need more
people to fund people like Danny and Susie to do the fundamental questions.
Yeah, so Susie, I think we haven't crossed to you. What would be on your wish
list then? Love looking at things at the single cell level. So I think a lot of
the time by taking it, you miss out a lot of these really exciting individuals.
You miss out a lot of these subpopulations that, like Danny was saying, are
really important. So for me, it would be single cell genomics in TB and also
single cell transcriptomics. So I know that's very difficult in bacteria, but
I'd love to see more of that. I think there's a lot of people who are now
preparing to try, not a lot, some people are preparing to try to handle that
bioinformatically because the difficulties are this culture bias that we require
if we are going to be doing things in the lab. So there's kind of two directions
that bioinformatics is going in TB. One is really trying to work more in the
clinical sample side and be able to do genetics and bioinformatics quick in an
easy way that's reproducible in the clinic by people who don't use
bioinformatics tools, which I'm sure everybody has problems in all different
pathogens with that. Then the other direction is how do we make sure we have the
entire genome? And how do we know that we're not just missing some sections,
which means more closed genomes, more long weed genomes and the proper
processing and DNA extraction for those. It's the same as it is in other
pathogens. It just feels like it's a little bit behind in some ways. So Connor,
yeah, you just mentioned bioinformatic tools for public health. Could you talk
about what's happening in that space and particularly genomic epi? Transmission
linking in tuberculosis has a very long and convoluted history. So you would
start off with insertion elements and then patterns of insertion elements in the
genome and then matching them between two different patients and seeing if it
has the same pattern, then you would say it's maybe in a transmission cluster.
The Spoligo typing that we talked about earlier with these CRISPR patterns, we
have a VNTR pattern based ones that are in there called MIRU-VNTR. It can't just
be called regular things. It always has to have its own mycobacteria in the
name. And now we're using the whole genome. In the whole genome, the standard
approach is to call SNPs against the H37RV reference genome, get the SNP
distance between two different strains, and then if they're less than five or 12
SNPs, you're saying that they're in some kind of transmission cluster. Some
people then have moved to, let's say, back a step, potentially back a step to
core genome MLSTs directly from the genome, mainly a group in Germany that are
working on that and trying to do that for circulating strains. The SNP distance
has become the primary way of looking at it. It works quite well to find out
what your circulating strains are, but not so well when you want to know exactly
who infected whom because of the problem of the slow mutation race. You could
have zero SNPs in a whole group of people. So in Rwanda, we have a transmission
cluster that has been there since the 90s, and it's 12 SNPs apart between most
of the different ones that are causing most of the MDR-TB. So it's very slow
mutations that are occurring in that time. We're moving more towards a Bayesian
approach of looking at transmission-based, or some of us are, not all, using
transmission approaches from Caroline Collin and Xavier Didele to try to say,
okay, well, we think there's these numbers in between, which take into account a
generation time and take into account an infectivity period, which can model
better the mycobacterium tuberculosis ones. But we indeed, again, going back to
what we said, is we need to be a little bit more specific on what is that
generation time? What is that latent or infectivity period to try to better
inform our genomic approaches for looking at transmission? We need that
circularity of, we can do better with transmission, but we need to know more
about the underlying way that it is transmitted to then come back up to it. But
in general, the SNP approach seems to work quite well, because we don't have a
combination, because we don't have plasmids that are going to mess up a lot of
those things. So working in enterics from my background, you're always at
loggerheads with some of the pre-genomic genotyping methods. So serotyping is
kind of, there's always like a love-hate relationship between things like
serotyping. How is things like spoligotyping for MTB? Is that generally
consistent with what people would see in a phylogeny? Or do you all have the
same kind of energetic enthusiasm about these methods that I do for other
organisms? Spoligotyping is the one that persists for longest. And that is
because it's less than a pound to do a spoligotype, and you can do 40 of them at
one time. So when you're in a low resource country, you can bang out those
spoligotypes really quickly and give you an idea of the lineages that are there.
It will not tell you that if it's a transmission cluster, and that's difficult
to try to get people to be on board with that, no matter how many papers I even
personally publish about that. So it'll tell you very well that the main
lineages that you have, lineages 1 to 9, and whether it's an animal one. But
beyond that, it won't tell you. You have a lot of convergent evolution within
those patterns. But again, there's the throw-up between accuracy and cost. When
you're just trying to look at what are the main things that are circulating,
spoligotype can tell you that. And what we're trying to get people to move
towards is, you know, spoligotype everything, and then the patterns that match,
maybe do something further on those ones. And if they don't match, maybe that's
interesting as well, but that's two separate research questions. You know, if
you have a new pattern, maybe we sequence that. And if we have things that are
clustering a lot, maybe we sequence some of those to see if they are a
transmission cluster. So there's a use for it, but it's not obviously going to
be the be-all and end-all. I think what a lot of bioinformatics needs to do is
marry those two together, of not say, don't do that, but say, there's a use for
that. But it is not everything that it was five or ten years ago, but it can
maybe point you in the right direction, like a serotype can maybe point you in
the right direction, but it's not going to tell you what we thought it did. I
mean, the issue with serotypes is sometimes you do have these edge cases where
it puts two things together that really shouldn't be together. And is that the
case with spoligotyping, or you don't really have that? Sometimes we'll put them
together, but it's not as, we don't use this the spoligotype clinically as much
as a serotype is done. It doesn't have this label of, now we treat in this way,
and it has these characteristics. But we discovered lineage eight because a
weird spoligotype came up in a routine thing. It just had none of the spacers
that anything we'd ever seen before. And my boss, Barbara DeYoung at the time,
she's just one of those people who has all the spoligotype patterns in her head.
And she went, that's not a spoligotype that we know. And then we just sequenced
it, and then it came up as a new lineage. So it has its use. That was just
because we were going through. TB, thankfully, doesn't have this means this, in
terms of a spoligotype pattern for most of them. What we are seeing more is that
things like, like Susie had said, you know, Bovis has intrinsic pyrazinamide
resistance. Kineti has that. And now we're working towards more a clinical
taxonomy of, can we better tell you from a label, from a lineage label, what
potentially is going to be the issue with the treatment. And that's something
that I'm watch this space on, hopefully. I mean, I guess that's a silver lining
there that you don't have that history, because so much of the serology and
genotyping is baked into government policy for other organisms. And that is
something you always have to keep fighting that they do have to wrap people on
the fingers for trying to find the same MLST type and saying, like, oh, there's
transmission between the cows and the farmhands. It's like, no, no, no, no, no,
no. And this is for E. coli.  makes no sense. Back in the day I used to work in
diagnostics the way we told if it was if it was bovis or or MTB is we'd have one
LJ slope with pyruvate in it and one without and it was as simple as that right
if it if it couldn't grow without pyruvate it was bovis and that's how you
change the treatment and that gets back to the kind of resource limited it could
come up with errors but it's cheap as anything right to say that's probably
bovis and that's not and that's all you really want at the end of the day no
definitely there's that definitely there's scope for for these methods and there
always will be that they're useful and they tell you something informative it it
becomes problematic when people tend to treat them as gospel and as kind of
saying like you have this then you must follow this it must behave in this way
it's like no it's it's it's a microbe it does what it wants you can't you can't
categorize it also it takes but it takes about a month for them to grow on those
slopes so we do need we do know we need the sequencing we need something before
that because by that point the patient's already been given pyrazinamide and
they've been on it for a month and you know so we need we need both we need we
need the genetics straight away which are fast and which will tell us what's
happening so any final comments from all of you anything you feel that has to be
put out there in the community they have to be aware of well i guess for me the
thing that we always say in the tb community that in the uk it's it's it's very
difficult to get funding for research into tb and you know with a lot of the
research funding being directed towards covid that we must not forget the other
pandemics that are still going on and that's not just tb because you know part
of the reason we we had such a difficult situation with covid was you know lots
of people who worked on coronaviruses actually lost their jobs because they
weren't considered interesting so i think if the if the covid crisis has taught
us anything it's that we need a broad brush of research and it's certainly no
time for complacency when it comes to mycobacterium tuberculosis and
tuberculosis that would be my final so to add on to that what's interesting is
that when the who released its top pathogen list mycobacterium tuberculosis
wasn't on the list and that's because it's actually a star at the bottom that
says tb is so far beyond all of these we thought it would just take up too much
of the list but the problem is that the funding agencies go it's not on the list
so we're told that it's not on the top pathogen list and we should focus on the
top pathogen list where the who says oh everyone knows and it's like not
everybody knows actually most people don't know that tb is killing as many as
the entire rest of the list combined almost that was the same with the amr list
as well wasn't it's exactly the same not on there it's not on the amr list
either yeah despite it being half a million a year despite it being the top amr
pathogen so yeah all right well that's all the time we have for today this is
part of our ongoing series to talk deeply about a particular microbe and today
we've been talking about mycobacterium tuberculosis i want to thank our guests
today doctors suzy danny and connor for joining us and we'll see you all next
time on the micro binfy podcast thank you so much for listening to us at home if
you like this podcast please subscribe and rate us on itunes spotify soundcloud
or the platform of your choice follow us on twitter at micro binfy and if you
don't like this podcast please don't do anything this podcast was recorded by
the microbial bioinformatics group the opinions expressed here are our own and
do not necessarily reflect the views of cdc or the quadram institute