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 is 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. The
fraternity of bioinformatics includes people with vastly different backgrounds
and expertise. You can have a strong computer science background or trained as a
bench microbiologist who later moves into bioinformatics. Today we're going to
talk about the advantages and disadvantages of these different backgrounds and
what people should be mindful of. To balance out this discussion, we have a
special guest today, Phil Ashton, who is a bioinformatician at Malawi Liverpool
Welcome Unit based in Blantyre, Malawi. Phil in particular moved into
bioinformatics post-PhD, and he will be representing the wet lab perspective. So
Phil, who are you and what do you do? Hi Lee, thanks for having me on, first of
all. So yeah, so my name is Philip Ashton, I'm a bioinformatician at the Malawi
Liverpool Welcome Institute, based here in Blantyre, in southern Malawi. I moved
to Malawi to work on Salmonella typhi and invasive non-typhoidal Salmonella. So
it was a bit of a return to the kind of home stomping grounds. I kind of worked
on Salmonella genomics when I was based at Public Health England previously. And
in between those two, I spent three and a half years in Vietnam working on
fungal pathogens and tuberculosis primarily, a few other things in between. And
so I've gone kind of from the public health bioinformatics perspective, I guess
similar to Lee, to well, similar to all of you post-COVID, but especially
similar to Lee, into the kind of academic role, just being a postdoc on a single
project for a couple of years. And then I was a kind of the lead
bioinformatician at OCRU for two years with a kind of remit to support
bioinformatics across the unit as a kind of micro, micro version of Andrew, I
guess. So then after two years there, my wife and I, my wife's a microbiologist
as well, moved to Malawi, where I'm working again on a kind of single organism
focus, Salmonella, lots of interesting studies going on, so I'm kind of just
leading on the genomics there and exploring options for kind of future career
independence. So prior to that, since we know you moved into bioinformatics
post-PhD, what, what was your background on the microsite? So I just did a
basic, my degree, undergraduate degree was just applied biology and actually,
you know, retrospectively I can see that, you know, maybe it was preordained,
but I had to make a bit of a fuss to like get onto a particular module called
genomics, proteomics and bioinformatics while I was an undergrad, you know, and
then learn about BLAST and things like that, that you do in undergrad courses on
bioinformatics. So I did that. And then maybe that was one factor that helped me
get a PhD. So genomics was in the title of my PhD, but what, what my supervisors
actually meant was AFLP. I did my PhD mostly in the wet lab with a little bit of
RNA sequencing, which again, I kind of had to make a little bit of a fuss to get
in. There was due to be some microarray work in there. Wow. That's a dodged a
bullet there, but basically as part of that, my PhD, there was this one chapter
on RNA sequencing. I was based at Public Health England and basically I didn't
have a clue what I was doing. So I went down to the bioinformatics group down in
the, in what feels like the basement down at Public Health England, and
basically bothered them until they let me use CLC bio to do some, some genomic
analysis. So that was my, my kind of background, zero command line or scripting
pretty much, you know, a little bit of R and things. A couple of times I'd been
kind of tried to learn Perl. Someone at one of the biophysicians gave me a, one
of the Perl, O'Reilly Perl books, but I got about two pages through one of
those. And then the kind of big change came from my first postdoc where I got a
job working with Tim Dorman. And then that was purely bioinformatics and kind of
in at the deep end. Yeah. Basically from there I've been more or less entirely
bioinformatically focused. I don't think I've touched the pipette in the last
eight years. So were you there for the very first wave of bioinformatics coming
into Public Health England? I wouldn't describe myself entirely as in the first
wave. I was probably in the second. I think I can comfortably say I was in the
second wave. So you know, Anthony Underwood and obviously John Green and Tim and
Steve Platt and everyone that had been there for a little while before me, but
yeah, I was there. I think the group, the bioinformatics group grew from yeah,
six or seven to 10 or 12 during my kind of time as a bioinformatician there. And
that was the time when they started just sequencing every salmonella coming
through the door, wasn't it? Yeah. So that was my kind of job really. So at
first I started with a postdoc working on E. coli 157. And then after that, I
kind of started as a, I was quite lucky, I guess in a way I had a permanent job
working as a bioinformatician supporting the salmonella whole genome sequencing.
Yeah. But when we made the shift from a kind of hodgepodge of different micro
microbiological methods to just using whole genome sequencing for everything. So
that was an incredibly exciting time and you know, really fulfilling period in
my career. You know, there's just something about when the phase, when
everything's beginning, that actually you have a lot of freedom, you know,
you're not so much constrained by how things have always been done. So you have
a lot of freedom and you have potentially a lot of influence at a relatively
junior stage in my career. So yeah, that was, that was a real kind of high point
for me. Yeah. I think in the public health space, people still point to that
change over in PHE and as a, as an example of what can be done with genomics in
public health, even now, like that's, that seems to be a very strong legacy
from, from what you all were setting up back then. Yeah. I certainly hope so.
And obviously, you know, the continuing work of Tim and lots of other people in
that, in that group. But yeah, just, you know, to talk about Salmonella for a
minute, because I know it's an interest of Nabil's at least, you know, one of
the big problems with public health, with sequencing for public health is that
realistically you're looking at about a month, you know, two weeks to a month
lag between the isolate being taken between, you know, the isolate being taken
and the sequencing results being generated. So if there's a bunch of people who
get sick at a restaurant because they've all had Salmonella food poisoning from
some dodgy eggs, then it's not incredibly useful necessarily to have sequencing
for that. But what it's, what sequencing has really enabled us to do is find
outbreaks with completely different kinds of epidemiology. You know, there's one
paper, there's one outbreak that's been published where there was these this
contaminated feeder mice that people feed to their reptiles, the snakes and
such. And lots of kids had got sick because you have the defrosted mice. And
looking back, we'd gone, the outbreak had been kind of dribbling along for at
least three or four years. And just kind of, that's the background against which
we identify these outbreaks in restaurants and such. So really, you know, with
sequencing, maybe it's still a bit useful for the restaurant type outbreaks and
you can include it in your outbreak definition and everything. But you know, the
background isolate, that's what I was, that's one of the things that I really
thought that sequencing could kind of completely change the paradigm of how we
deal with Salmonella in high income countries. So now we have about 200,000
Salmonella sequence. So, you know, when do you have enough sequenced? Yeah, when
you're doing it prospectively, when you're doing it for a public health
intervention, then, you know, never really, you know, you still need those, it
depends on your application, right? If you're, if you're doing research purely,
you know, if you're doing evolutionary genetics or population genetics, then,
you know, we probably have, if we stop now, then, you know, probably no one
would cry into their evolutionary genetic software about that. But, you know, if
you're doing prospective sequencing in order to inform public health action.
then hopefully that's just gonna be the norm for the next however many years
until the next generation of typing method comes along. Can I ask a silly
question? From much earlier, you said that you started off in the basement,
which I thought was funny when Andrew brought it up on an earlier, much earlier
conversation. Is there just like a trend of basements back then? Or did you guys
meet each other through the basement connection or what's going on with
basements? There's a secret underground city of bioinformaticians in the UK.
Yes. Yeah. It's like Milton Keynes, but just underground. So that's just a
really, basically a public health England. They kept the bioinformaticians in
this kind of slightly smelly office that was kind of tucked away at the back of
the building. Wasn't actually a basement. It just had like an overhang outside
and then a backup generator. So there was basically zero natural light and it
felt like a basement and it smelled like a basement. So yeah, it just kind of
jokingly referred to it as the basement. One more follow-up question to that is
like, was there a time in that history where you guys moved out of the basement?
Did you get to move out of there? So actually one thing that I thought was very
interesting or is that so me and Tim were based in, we weren't in the basement.
We weren't alongside all the other bioinformaticians. We were just the two, we
were embedded within kind of the gastro bacteria unit. And this is the kind of,
people will go back and forth on what kind of model they want. Do they want a
hub and spoke model for their bioinformaticians, which is kind of what we had at
PHE or at least me and Tim were a spoke and there's the big hub of central
bioinformatics. Or do they want everything to be centralized? And personally, I
think that the hub and spoke model is superior just because if you want these
tests and these assays that we're developing, bioinformatic assay to be useful,
then really it's helpful if the people developing it are in daily contact with
the end users, with the microbiologists and the epidemiologists who will use
those outputs rather than kind of being in the basement with the other
biophysicians. I see the other side of the coin as well in that I once spent at
least a month, a horrible month of my life writing a Python application that
wrote an XML file to upload these 10,000 genomes a year to the NCBI pathogens
portal. And then three months after that, discovered that one of the
biophysicians downstairs had just spent the last month writing exactly the same
thing. So, you know, it's just, that was a bit frustrating because I think I'd
rather have, I'd rather have that communication with the microbiologists and
epidemiologists. So given your background, what was your first project with
bioinformatics and then specifically, and then what did you find challenging
moving into that space coming from a micro, like a pure bench micro background?
Kind of distinctly remember my first day in this new job looking at, it was E.
coli or 157 genomes. And Tim said, oh, you know, just SSH into this server. And
I was like, what does that mean? Yeah, I just had zero clue basically about the
command line. And I, you know, when I'm training people now, I kind of try, I
really try to kind of remember back to how that was and how that felt in order
to kind of boost my empathy with the person I'm training, which sometimes works
and sometimes doesn't. But I think that actually the best people to do training
of complete rookies, maybe are people who've learned, you know, within the last
two or three years, because they can still empathize with finding the command
line confusing and not, you know, not kind of navigating a remote server kind of
like second nature by second nature, which is that most people who've been doing
it for five, six, seven, eight years. So that was kind of definitely one aspect
that I found intimidating. But, you know, at that time I just had, I had kind
of, and Tim was very like generous and I had like enough space to kind of learn
all of that stuff on the job. And then the other aspect was kind of the Python's
coding. And I just remember, you know, the first, I was writing this pipeline,
which is just a series of commands, shell of a OS dot system calls in Python.
And I know I'm not supposed to use those, but this was eight years ago. And then
going home and that thrill of, you know, the next day logging on and seeing it
all run for the first time. And, you know, all of these E. coli genomes had been
analyzed and, you know, how kind of satisfying that was. And then obviously, you
know, that was absolutely hideous code, would be ashamed to have now, but, you
know, that's just kind of, it's good to remember where, where I guess we all
started at one time. So you mentioned that the challenges was the basic sort of
technical stuff around SSH and just basic programming. Was there any major
concepts that took you quite some time to understand that people like Tim would
just whiz right through and you're like, just conceptually, like, I'm not,
you're not there with, you're not able to follow. Yeah. Like Tim was and is
probably a better coder than me, but I guess, you know, you pick it up in kind
of two ways, obviously one by just like learning and bashing your head against
it, but also by being in that group of, you know, five, six, growing to like 10,
12 other biophysicians, including some really great, really great people. And we
used to have these Friday, like coffee morning, like round table things, people
would just go around and say what they were working on. And that just helps you
pick up like the lingo, right? The, all the terminology is like, okay, you know,
gamers and, you know, like hash. So it was like, you know, KH had come out and I
was really excited about that. And, you know, just kind of, you know, but
hearing more experienced people talk about it as well, just kind of let you
advance so quickly, so much more quickly than you could by yourself. And, you
know, that's where also like Twitter and are we allowed to mention the S-L-A-C-K
word? Yeah. Yeah. Yeah, Twitter and, you know, the Slack group, that's where
they still provide a really important function for me now. Because, you know,
I'm a kind of, there was a couple of other bioinformaticians in Vietnam. I think
there's a few people who do genomics here in Malawi, but probably no one else
who'd described himself as a bioinformatician. So, you know, maybe even in the
whole country in Malawi. You're the only bioinformatician in the entire country.
Well, I mean, I haven't asked everyone yet, but I'm not sure where they would be
if they are here, because I'm not sure there's another one at MLW where I'm
currently based. And I think we're probably the biggest kind of research center
in Malawi, but don't quote me on that. So your lab meetings must be very
interesting. In a way, you know, bioinformaticians, even when you work in the
same building as biophysicians, then often you'll interact with them online
anyway. So in a way I get to benefit from, you know, the things I wouldn't see
if I was in, you know, a bioinformatics group in the UK, which is kind of the
clinical and epidemiological sides of things, of infectious diseases in Malawi.
And I can still get my, some at least, it's not the same as being in a
bioinformatics group, but I can still get some of the same interactions on
Twitter and on Slack. So how did you find going from like a public health body
where you had access to basically any sample you wanted, all the metadata, all
the PPI, to say academia where you don't have any of that and you know, you
struggle to get samples and you have probably no patient information whatsoever.
So I was really lucky actually. Well, lucky and, you know, kind of chose the
projects that I did quite deliberately. In the, at least for the fungal work
that I moved from PHE to do, the patients had been in clinical trials. So the
metadata associated with them blew what was available to us at Public Health
England out of the water. You know, you knew that their blood pressure, you had
like a thousand blood pressure measurements for each patient and, you know, you
knew how sick they were when they enrolled, how did they respond to treatments
for all kind of 700 people in the trial. So that's really nice. And I'd highly
recommend any biophysicians to go along to your local medical school, see who's
doing clinical trials for infectious diseases and suggest to them that they, you
could put in an application to do some sequencing together because they're so
expensive.  to do, they're so well-characterized, you know, just having that
sequencing data on just lets you, it's a kind of, you know, what they like you
to say in funding applications, it's highly leveraged. You know, they spent two
million quid on that clinical trial, so, you know, you may as well give me 100
grand to do some sequencing. You keep answering a lot of the questions that I
have, actually. I wanted to ask you what your support is like, and a lot of what
you're saying is that you have to go online, basically, on Twitter, on Slack, on
wherever, to kind of get support for what you're doing, right? Is your blog kind
of like a therapeutic thing, too, in that regard? No, not really. Blog was
mostly, definitely started out, you know, before I was really very good at
writing papers. I wanted a kind of, something, a forum that would give me
writing experience that was easier, you know, had a lower bar, lower activation
energy than the whole paper-writing jamboree. And it's just something, there's
just something about, you know, sharing, knowing that something's gonna be
public, that, like, less so now, because I'm more confident in my work, but, you
know, when you're a real rookie, you know, rather than just doing some aspect of
the project and just, you know, writing it up in your lab notebook or whatever,
if you turn it into a blog post, then you have that extra bit of peer pressure
in the back of your mind that makes you double-check everything and kind of, you
know, push the project to kind of a neat conclusion, rather than leaving things
hanging. So that was the idea behind the blog originally, and now I just kind of
occasionally use it to post, you know, how-tos and little bits and bobs, really.
Do you see yourself as an academic or maybe a professional blind practitioner?
Hmm, yeah. I mean, depends what kind of job I'm applying for, right? Well, you
know, academic, as in, you know, applying for funding and teaching and all that
kind of jazz, versus maybe more professional services. Yeah, no, I think I do
see myself more on the academic side of things. Not sure the rest of the world
agrees with me, but. So you want to keep your options open, that's what you're
saying? But yeah, I would definitely go back and work at Public Health England
again, but I'm not, and you know, a specific institute, or working at an
institute like Quadram or something that has quite a tight focus on microbes,
but I wouldn't be very keen to work in a kind of general university core
bioinformatics support kind of role, really. I'd probably rather leave
bioinformatics and go and do something else. And no, being more serious, yeah,
you know, I think it would be nice to be back in a bioinformatics group. You
know, I often think of how good it would be to spend even just like one month or
a couple of weeks a year, you know, well, maybe not Sanger anymore, but, you
know, Quadram or PHE or, you know, somewhere with a real kind of active core of
biophysicians. So like a critical mass? Critical mass, exactly. Yeah, just, you
know, you just learn. So much just by osmosis from being in those groups that I
do miss. Well, we'd love to have you. And if anyone else who's as experienced as
you wants to come and work in Quadram, you know, we'd have you as well. Everyone
is always welcome. Obviously not now with COVID, but, you know, in normal times.
Knowing what you know now, what do you wish you knew like before when you
started? What would you go back in time and tell your past self regarding
bioinformatics? Learn SnakeMate. How about learn how to use Galaxy and also
learn how to program in Perl? Yeah, the Perl one for sure. Yeah, I think that if
I was to go back in time and give myself one piece of advice, it would almost
definitely be really engaged with a workflow language. You know, a lot of what I
was doing at PHE was writing scientific workflows. And I just, you know,
bootstrapped them together with Python. Whereas more directly using, you know,
SnakeMate or NextFlow when it became available. I think that would have led to
some great productivity gains. So on a technical level, I think learn SnakeMate.
Today, if I was telling someone just getting started, they have to be quite
strong with programming. You can't do anything reasonable with just mucking
around in like a CLC workbench or clicking around. You need to start thinking
programming, workflow management, have to operate at scale. I mean, yeah, if you
want to work at stuff through Galaxy, then you can also push stuff through that,
but you'd have to want to learn the API, BioBlend API as well to really get the
most out of it. Actually, I would say interacting as much as possible with
biologists is probably the key for me because I came from a computer science
background. And of course, you know, you learn how to program and all of that,
but actually the biology and the in-depth understanding and the quirks of all
these different bugs, you know, it's something you really need to understand
from an expert. Yeah, I think that's what Phil had from the other side, really,
which was interesting was that knowledge. I get that I'm putting words in your
mouth, but the knowledge exchange was crucial for you to get to where you are.
Yeah, I think that- I mean, you're going micro to computer and then Andrew also
saying the same, but in reverse. I mean, that seems really critical. And it's
sad that you still have pet biopharmacicians who don't get that support. Yeah, I
think being able to speak lab has definitely been very helpful in multiple parts
of my career. I would even add on to Andrew's answer that like, if you're in the
public health arena, not only talk to the biologists, but talk to
epidemiologists or whoever else is in your area so you can get a more global
understanding of what you're doing. Because I also agree, I wish that I learned
more programming early on, but coming in with biology knowledge and public
health knowledge is just so priceless. So thanks, Phil, for joining us today.
We've gained a much better appreciation of the lone bioinformatician. I really
enjoyed learning more about Phil and I hope you did too. Please join us next
time where we dive deeper into Phil's story. Thank you all so much for listening
to us at home. If you like this podcast, please subscribe and like us on iTunes,
Spotify, SoundCloud, or the platform of your choice. And if you don't like this
podcast, please don't do anything. This podcast was recorded by the Microbial
Bioinformatics Group and edited by Nick Waters. The opinions expressed here are
our own and do not necessarily reflect the views of CDC or the Quadrant
Institute.