This is a machine transcription and therefore it may contain inaccuracies, errors, or mispronunciations. Notice an error you think needs changing? Please contact the Bitesize Bio team using this form: https://bit.ly/bsbtranscriptions

Welcome to the Microscopists,

a bite-sized bio podcast hosted by Peter Oto,

sponsored by Zeiss Microscopy. Today on the Microscopists.

Today on the microscopist,

I'm joined by members of the network of the European bio image analysts.

New bias coming up,

co muira shares his advice about go freelance with image analysis.

I think I, I don't really recommend to know more people when, uh, so I,

I'm bit weird person. I, I confess myself. I mean,

I'm very optimistic. So, uh,

so the first thing that he become a freelancer is,

I think you have to be optimistic.

Robert Haa explains why he prefers to invite end users in on the coding

process.

Why? Nowadays what we, what we do more is like,

instead of providing a script to people and then hoping that they use it in the

right way, that's another topic. Um, we write the script together with them,

but they learn something. Um, and then also the next,

the next project our collaborators want to do,

they could potentially do it without us because they acquire this skill.

And El nurse FLI chairs had her new bias image analysis score helped her through

a tough period in her PhD.

In the middle of my PhD, I was actually, um, stuck.

I was doing a lot of microscope. I started with Ted, so quite deep. But, um,

then I couldn't analyze so well, I mean,

everything was manual and time consuming.

So that's when I got to hear about new bias

All in this episode of The Microscopist.

Hi, I'm Peter Oton from University of York,

and welcome to this episode of The Microscopist. Today we have three, not one,

not two, but three guests. So we have Elna from Turkey,

we have Robert Haa from Germany, Dresden,

and we have Komura who's freelance, Koto.

Where are you living at the moment?

I'm in Japan now.

Yeah. Right. So different time zone again. So, because I, I remember, yeah,

yeah.

Uh,

Oh, back in possibly 2002, 2003.

When you were at E M B L

2000, I think. Uh, so were you in, uh, MIE in Paris?

I was, no, I, my first one was, uh, Gothenburg.

Okay. Then maybe later than, but I, I,

I kind of very strong impression I have about, about Peter Is,

uh, this the one that you organized Elmi in 2000,

when was it?

Uh, 2000 6, 0 7.

Yeah. Yeah. So, uh, you had a, you know,

quite a gorgeous lineup of beers from, uh,

local beers. And then, uh, there was a hush puppy.

I think so, uh,

somehow I had an impression that all these beers,

uh, they have a name after docks. That was it.

Yo Shater. And, uh,

Yeah, there was Yorkshire Terrier there. Yeah.

Yeah. That, that was it. Uh, so, uh,

that kind of made me strong impression about you,

because you were proudly presenting all this beer on the stage.

I can't believe your impressions of me are not about size, but about beer.

I think that's,

No, no. I, I already said it was the first impression, you know,

so afterward you became like a hero. I know,

but the first thing I recognize about you was, uh,

strongly associated with this beer, you know? Yeah.

That's cool. And I'm gonna bring this back up later.

'cause I might have one or two photos from that meeting. Uh, but I,

I'll come to that in a little bit. I,

I, I, I hope it's not too bad. I mean,

I, they're,

and I'm sure Robert and all knows are gonna really enjoy seeing them. Uh,

right. So we should, we should get onto the topic of today.

So today's slightly different. We're gonna focus on new bias. Uh, so actually,

who would like to explain what new bias actually stands for to start with? Uh,

shall i, I el

I can

Go,

I can do that. Yeah. Yeah.

So we named Novi New bs. I mean, uh, so it's,

uh, the network of European bio image analysts.

And then, uh, it comes with the history of, uh,

how we started naming courses. So that, uh,

the initial course that was in, uh, the m o,

uh, courses, um, this is called,

uh, I, I think it was the na, the name was, uh, um,

the master course for Bio Image Data Analysis. And then,

uh, I, we started to abbreviate as a bias bio major analysis.

Uh, and then, uh, and then, so there was, uh,

um, so that was 2012, I think that we named it.

And then after this, um, we used this bias often for, uh,

as a kind of a, um, convention to just name things,

bias. Bias. And then, uh,

there was a European biomedical Analysis symposium,

which was in Barcelona, and this was called U Bias. So,

uh, EU Bias. And then, um,

this is because, uh, we kind of, uh,

integrated the original course contents and then make it shorter a bit,

but, um, with more people in a sense. And this is U B s,

then at this time, we didn't have money,

and then we started looking for getting money. And then now we got money from,

uh, EU cost cost eu. And then, uh,

for this project we added n in the beginning of EU

bias, and then there's no bs. And then, uh,

many people think that it somehow comes from a German new, uh, which is no,

uh, but it's not really, uh, true.

We try to edit, make some additions.

And then now we are trying to go for Sobi, which is still,

uh, um, we haven't the hypothetical name still,

but the project name is Sobi, which is Society for Biomedical Analysis.

So, uh, we kind of, um,

stopped eating more rather than we chopped off the beginning. And then,

uh, now it's Sobi.

So the reason for going for Sobi is to bring in the rest of the world,

is that correct?

Exactly. Yeah.

So that's now the led by robot who is sitting

here now.

And it was actually, it was actually a hot discussion if, uh, how,

how we can call it, I can't,

I can't re I can't recall all the names we had in mind,

but there was also the network of bio image analyst, no, by us.

And then there was a Globe buyer. And then we were thinking about,

of course, you know how Zoom meetings go,

you have something like 50 people in a Zoom meeting, and of course,

we didn't find the content. So, um, the final name is not decided.

It's a working title, um, Z buyer.

And I would say that we are now planning to fund a society to

become an actual, um, body with like, um, a legal entity,

uh, with, uh, cotton Penta and these kind of things,

so that we can also acquire funding ourselves and do not have to rely on host

organizations. And when we, when we found this thing,

then we have to have a name.

So we can now go through the process of writing down our, our,

our internal guidelines, our rules, um, our, uh, governance model.

And with that, we also have to figure out how we want to call ourselves.

Ideally, it is something that everybody feels welcome.

Who analyzes images from microscopy biomedical contact?

So who, who were the founding member? Who,

whose idea was it to start this in the first instance?

Who were the original sort of figureheads to kick this off?

The first person I had ever talked to was Julian Elli, who had this idea.

But Cota you may know better. How,

So, um, so this was, um, do you, you know, that,

so it actually comes from Elmi. Um, so,

uh, uh,

it was already after we started this, uh, um,

a bias course. I mean, uh, at this, for this course, Julian was not involved.

But then, um, I think there was a meet Elmi meeting in, uh,

Bordeaux in, so there was a two times Bodea meeting,

and then the one that was earlier, I think,

which was very bad weather or something. Do you remember this? Uh, yeah,

I've been to the, I remember Bordeaux very well. Yeah. So that was the next one.

Do you, so, so we, when was this?

So I forgot about the 2000, I think

14 or 14 or something like this. Right, okay. And then, uh,

so that, uh, I wasn't there for, I think,

uh, um, at this point I have a lot of family problem. I mean,

not problem, but, uh, I was busy with family things, and then, uh,

I wasn't really, I stopped going to MIE to save my time.

And then, uh, um, but then,

so I always hear a lot of this, uh, what happened in Mie at this,

around that time, the people just come to me and say, oh,

there was this and that, and now me and so on. And then this time, so, um,

actually Julian called me directly after he went

back to Barcelona. I think there was a call and then, uh, saying that,

um, there was some, um, many people asked, um,

we have to start some kind of, uh, um,

network with bio major analysis. And then, uh,

so, and then Julia said that, um,

so I think you have some idea about this, so that, uh, yeah, yeah.

And then so, um, the initial thing that, um,

I proposed to do was to write a short, um,

um, text about the plan of how this network should be, and then, uh,

how the meeting of this analysis should be. And then, and then,

so, um,

I wrote the draft of, uh, the kind of this, uh,

skeleton of how the meeting should be, and then this should be combined.

So there was already this idea of, uh, making a,

a multiplex meeting of courses and the symposium.

And also, so, uh, this hackathon actually came after,

but combining this, uh, symposium and courses. So that wasn't like a,

so this is, uh, um, the text that, um, I wrote. And then,

then the main point was this, but, uh, um,

while there are many meetings about the software, yeah. So at this time,

actually, so there was a still image, a symposium,

which was held, um,

every year between Europe and United States. But,

uh, so it's alternate the, uh, the place. And then, uh,

so I was there several times for this image meeting.

And what I felt was that they always only talk about software. Yeah.

So that they don't talk about analysis. So did, that's a kind of a,

for me, it was a different because of,

I went to image meeting and I was kind of, uh, became a bit disappointed.

So I was very excited to meet like, uh, Wayne Ra Band and all the plugin stars.

But then, uh, they, but they only talk about this individual,

how plugin works and how to use them,

but never really talk about the analysis itself,

which is what I'm interested in. So, uh,

so I was kind of half satisfied, 50%, but, uh,

it's not full excitement. It's like a 50% meeting actually,

like, uh, um, um, and, uh,

ish. So those, those, you know, or you Johanne Lin or,

uh, Abad Kaon who made the, the, the fi Yeah,

the meeting name was exciting, and hearing how they made it was amazing.

But, so I had this missing part,

which is about image analysis itself,

which I always hear some of them in like the Army meeting or, uh,

in, uh, cell biology meetings where you actually have some biological problem.

And then trying to tell how they solve using image analysis are the

kind of major tool for, uh, measuring something. Now,

I wanted to have a meeting where really physically only talk about this,

what I'm interested in, so that the,

the major message of this text was, okay,

let's do some courses that are actually centered on these bio biological

image analysis. Yeah.

Not only about flaggings and how the software architecture is,

and then how the algorithm,

it's rather that more centered on how we measure various

aspects. So, so based on this idea,

Julian Cobe and Sebastian Toshi made a lot of, um,

feedbacks on this draft. Okay. So we can do this and do that. And then, so one,

one thing that, um, came up during this, uh, um, discussion,

which was very informal. We, so I, I don't know how we did it,

even though at this time we didn't have Zoom.

So maybe it was a phone conversation because Julian and, uh,

Sebastian are sitting together. Mm-hmm. And then, um,

so, so like, uh, the idea of this, uh,

bio image information index, uh,

was, came in there while we're discussing. So, uh,

so we need to somehow organize tools, you know, so the,

the one important message, we had another one besides this, uh, um,

um, to, uh,

explore biomedical analysis as a biological, um, interest.

So that's number one. But second is that the second message that, um,

while many people say that, um, um, so if you see those, uh,

plugin papers, it's always that there is no tool doing this. Yeah.

So we have a very, uh,

or it can be that we have a more faster processing or more precise processing,

or we have a more functional, um,

with a more functionality and so on. But I think so,

so at this point, so what we thought was that no, no, no,

there are too many tools. Uh, it's actually,

so almost on, like, uh,

every day we see some new tools appearing, and then, uh, so, uh,

it's rather that we want to know what is already there, right?

And then so what is already there, and then also that what they can do.

Uh, so, uh, yeah.

So Cota, how long did it put, how long, roughly how long? One year, two years,

three years to put this together and get the funding and start it. How, how,

how long did it take? The total effort?

So we, so it, so the text is in a GitHub, so it's called, uh, um,

um,

But from the time that Julian called, after that ELMI meeting,

to the point of actually having the first new bias meeting where people came

together and talk about how it's gonna move forward and everything,

once you got funding, how long did that take? Was it one year? Two years,

three years?

That's okay. So, so I think, so,

so that was a proposal. Then we started meeting, uh,

which actually was U B s. Yeah. So

You started before the funding started, you,

you were doing it even before the funding anyway, then the funding. So

Yeah, yeah. But, but, but UB USS is, uh,

partially funded by Euro, by Imaging, actually,

I think. Uh, so because, um, um,

Jason Swedlow was a big supporter. He actually,

Jason talked to Julian and that, so several people talked to him,

but, um, that, so Kota should do something together with you or something.

So that's how

Kicked off at the start. So I've gotta, just, just to move this on a little bit,

I've got to ask, when did Robert, Robert, when did you come into this?

And then Alan know, I'm gonna come to you and ask exactly the same question.

When did you come into this?

I'm actually also looking forward to learn when El exactly joined,

because I think, I don't know this story yet. Um, so I, I, I'm probably,

I joined in 2017 in Lisbon as a student. Um,

I was in the bio analysis training school, and Julian wasn't den before,

told us about it, and one of us should come.

So that's why I came there as a student.

And I was completely amazed because until this point,

I was sitting at the co facility engineer in a co facility, and I was,

I felt pretty isolated.

And then I was sitting with like 25 other pretty isolated corpus,

actually in a room together. Um, and it was like, wow,

there's other people who have exactly the same problem I have. And it was like,

super amazing. Um, and I wasn't,

I think a year later I was trainer on one of those schools,

and another year later I was then a scientific organizer of E-school.

And then another year later I started with others together writing proposals for

extended funding. Um, here, one important lesson in this, in this August,

how to came things together, context, the grant we have now, the money, which,

which we can spend now for hiring a community manager who will start in a month,

by the way, um, we wrote three proposals to get that.

So we had like multiple attempts that we had to do it again and again,

and not give up and do it again. And eventually we got this money. So,

but it was, it was a two year process, pretty much the entire pandemic.

We had two meetings again and again,

and then writing a proposal for this one and writing a proposal for that one.

And eventually we got some money to hire this community manager. But Elena,

how did you join us? I really would like to

Please el When, when did you join you guys? You, you are the youngest one here,

I think.

Yeah. I consider myself a second or third generation here,

because Scott was there from the beginning.

And I think at the point where you had the element meeting and everything,

I wasn't even in a, a bioimaging context. Um,

I was an electrical engineer actually designing PCBs and, you know,

programming with c plus plus microcontrollers and stuff like that. Uh,

so, uh, yeah, but in 2012, I came, uh,

to Finland and I did a master's in biomedical imaging,

and that's where I got to know microscopes and how cool they are.

And that's when it all started for me. Um,

then I continued to do a PhD, also indoor Finland. And, um,

in the middle of my PhD, I was actually, um, stuck.

I was doing a lot of microscope. I started with so quite deep, but,

um, then I couldn't analyze so well, I mean,

everything was manual and time consuming.

So that's when I got to hear about new bias.

And my first new bias, uh, encounter was in 2019,

um, when I went to, I think it was Luxembourg for, um,

image analyst school. So I was a trainee there. And,

uh, there were mainly also programmers and hardcore people. Also.

Robert was teaching there with the, with his Post-It notes,

and I, COTA and Julian and others. And I really, really loved it. So I was like,

yeah, I, I have to just keep pushing and keep learning more.

Um, the next year, uh, in 2020, I was, uh,

organizing and, uh, teaching in already in, um,

early career school in,

and that was when Covid was starting,

and they weren't sure what it is and what's happening.

So the moment that I landed back in Finland,

all the borders closed, and then Covid started and the pandemic started.

And so, um, we continued,

I was helping organizing the new Bias Academy at home, um,

webinars at that point.

And I was very active then with new bias and continued, uh,

connecting with people and meeting and, uh,

volunteering to participate and help as much as I could. And,

um, at the same time,

I also got a job in 2021 and moved to Helsinki to

biomedical imaging unit. So that was actually pretty good,

because then I also needed to improve still my own image analysis knowledge.

I had to help users with that as well. And microscopy.

Um, so yeah, when there was the fund funding calls and I helped,

and finally we got, uh, new, uh, the CDI funding.

So I was very excited to be part of that now.

So actually of, of all of this covid for all the negatives,

it has created new opportunities and enabled you probably given you a little bit

more time in those early stages to actually put the,

to get together better as a network Exactly.

To look at new ways of working and to put those proposals together, uh,

which is really cool. And it's interesting to how you got into it. Uh,

and Robert, I presume similarly, how did you get, what was your first degree?

Maybe I should ask that to all of you. So, COTA, what was your first degree?

Um, so I, my, my undergraduate is, um,

so this is called, uh, bachelor of Liberal Arts.

Bachelor

Liberal

Arts. Liberal Arts,

Liberal Arts,

Dr. Kirk, liberal Arts.

Liberal Arts is, uh, I mean, uh, it's, uh, um,

so you know liberal arts, right? No, it's a liberal arts college. So, uh,

so I studied, um, biology and philosophy.

So, so how to, I, I, I'll come back to this because I'm go, go,

just go around the table to find out what everyone did. Is there first, uh,

eleena your, your undergraduate electronics?

Electronics, yes. Electrical and electronics engineering. Yeah.

Love it. And Robert, so

I, I did it,

I did a little bit longer way of education before achieving my master's in

Germany, called the Second Way of Education.

You first do a formal training and a job and a thing,

and then you enter university when you are much older.

So I am a software developer, programmer by training. And then afterwards, I,

uh, did a master's in, um, uh, computer science, how,

you know, I wanted to do something different. Um, and then I did, uh,

at the medical faculty at the university in western, a PhD in, oh,

what's the technical, a technical position or something like that.

So it's medicine, some form of medicine,

but obviously from the technical perspective.

So none of your careers are,

or maybe coach to an extent are directly into life sciences to

start with. Uh, and coach, you, you were so coach, I,

I first got to know you probably 2003, 2002, 2003, summering,

uh, no, not summering, uh, Guttenberg. Uh, when you started at E M B L,

uh, or not long after you started at E M B L, what drew you to E M B L?

Just briefly. What, what, what was the attraction? Why, why that role?

Yeah, yeah, yeah, yeah. So, so it's actually, um,

so you know, y

Yeah, of course I know Y

And, uh, Tim Oman, you know them, uh, uh,

heroes. Yeah.

So they are from the same lab in Munich where I was.

So there was a very small lab is actually only,

only us working on, uh, mold dti.

Yeah. And then, uh, so we were intensively using, um,

we, we were making micro, I mean,

we were converting microscopes using microscopes con. So at this time,

confocal was very fancy. And then,

so YZ moved to Heidelberg who, and then, uh,

in, I think 1998 or something. And then,

uh, so he was working in a different lab, but initially,

but then he started working with Rhino Kolk in some,

something in between 2 19 98 and, uh,

2001, somewhere between, and then be,

because then envi of started the world first

advanced light microscope facility in the world. So this was the, the first one,

the facility, yep. And then facility concept came from, uh, this, uh,

inside envelope. And then RiNo was responsible for making it. And then yes,

actually, um, did a lot of these practical things.

And on the way he called Timo in Munich, uh,

because he was, uh, about to finish his PhD and said, Timo, could you help me?

And then Timo moved to Heidelberg from Munich. Then after this,

they were saying that quota, you should come as well, and so on. But then, uh,

um, I was not really, um, I, I wanted to keep on doing some, uh,

project that was working on a PhD and then do it in my postdoc time. But then,

um, the years told me that, oh, quota,

there's a very good three D time lapse microscope that you can do in Muni. Uh,

you can use it. And then, so I,

I drove my old golf too with, uh, samples and equipments,

uh,

some for special imaging to Berg.

And then, uh, so Timo said that, oh, I have an extra room,

so I just put my things there and then slept for one week, stay there,

do some experiment.

Then I said that up and I was already, you know, um,

very welcoming me to do some experiments with these facilities things. But then,

uh, so I needed another week or something, and I said, ah,

so I need another week. Then, um, writer said, no, no problem.

You can stay as as much as you want. Then that kind of, um,

very ambiguous start. And then I started to stay in ambo. And then,

so at some point I got some money from Japan to support my, uh,

research in Ebo. And then, uh, and then I started to stay in Heisenberg,

but a very long time, I was just living in this, uh, um, T flat,

uh, uh, not really determined to stay in Heidelberg or not,

but eventually it became really serious, uh, to be in, uh,

work as an BOL postoc. And then eventually was a, from 2005,

I became the image analysis expert.

So that was the way it went.

So that's how you got into that. And then, then the image analysis started, uh,

Elna, how, how did you move? So you went to your, you went through, you went to,

but why image analysis? Why not stay with the,

the imaging or the scientific question? Why move to the analysis?

It was from the user point of view, I was myself stuck. Um,

I had lots of images. I,

I was having so much fun spending hours, um, you know,

in the dark room on and looking at my samples under microscopes and just taking

pictures.

And I was getting to the point that I was taking pictures for others as

well. And, uh, not just with one modality, but with multiple, I mean,

I've got my hands on em, uh, confocal,

super resolution light sheet, and then a f m and all kinds of stuff.

So I needed to then deliver more. Um,

and I needed to kind of get numbers out of these pretty

pictures that I was getting, and I didn't know how to do that. So, um,

that's when I realized that I think it's time for me to,

to move on. Like, microscopy is great. I love it.

That's why I chose a job in a core facility. And, um, I, I really,

really enjoy doing it, but I don't think it's enough. Um,

you have to then, then move on from it as well. At some point, it's, uh,

it becomes your comfort zone, and the learning curve gets flat,

and then you get bored.

And I'm one of those people that when I'm get getting bored, then I,

I'm not motivated to, to do really efficiently my job anymore. So,

so image analysis is something that I'm not comfortable doing,

and it's really, really like hard for me to, um, uh,

get grasped with when it comes to all these new techniques that keep coming with

this super fast speed. Like when I started machine learning and deep learning,

okay, it was there, but n normal, I mean, uh,

every day biologists wouldn't need it necessarily for the analysis,

but nowadays it's coming more and more. So anyways, I find it really exciting.

It keeps me on my toes and, uh, I love it. And I,

I also like, um, staying in the image analysis field because of the people.

I love the community, and I'm a sociable person. I like, uh,

the communications aspect of it. So that's why, um, it kind of, it,

it attracts me because people are so nice and they're very helpful and they're

very open, and you are talking about fair and open and reproducible,

and they're, they're all down by Alex. So yeah, it's been a,

a really nice journey.

It's a good area to be in at the moment. Uh, which,

which we'll come to later to, it hasn't always been the case, as in,

and as Robert, you mentioned earlier,

you had to write three grants to get funding to start with, so, well, no, it's,

uh, I wouldn't say it's the emerging area anymore.

I'd say it's an area that is very much here, but getting funding has now become,

went from being very difficult to now being just as difficult as any other

proposal in the area, but at least that's better than what it was. Robert,

how did you migrate into life sciences even and into,

and why image analysis? Well, you're become science,

but how did you hook into the wanting to work with life sciences and solving the

help, help and develop solving problems? It

Started with the data. So when I was, when I, when I was a student,

when I started computer science, you know,

you learn pretty much everything about, uh, computer science.

And there was one thing which was very fascinating for me, uh, which was, uh,

three dimensional image data. Um, because there's like plenty of algorithms,

computer vision, detecting trees and stuff. This,

this research field is decades old. Um,

but when you think of three dimensional image data, there is much less.

And the big question is,

which or how can we translate algorithms for two dimensional data into a three

dimensional space? And it's like a big data problem. And when I started that,

it was example, like 2008 and big data was something like half a gigabyte

was really challenging to process this kind of data on the computer back in the

days. And this is how I came into a research center.

I can research center here in Ton where I was for multiple years,

the only computer scientists who was working there first as a student assistant

later as a PhD student.

So I was helping everybody in this institute with their image analysis very

naturally, because, you know, I was one of the people who could program,

there were also some physicists who were doing similar things.

So there was also a small community,

but I was the only software engineer slash computer scientist.

And then when I finished my PhD, that was in the neighbor building,

actually an open position for a bio image analyst. Uh,

I think they called it a bio image information back in the days, um,

in a core facility. So I immediately, after my PhD went into this position,

and it also became permanent at some point. So I had a permanent position in,

uh, actually a company which was working within a research institute.

And now comes the fun fact. Um, I learned, again, I,

I mentioned it earlier in my first new bio school,

I learned that there are more people who have similar problems, um,

and there are multiple approaches for solving these problems.

So I did not agree with how the problems were approached in the, in the,

in the department I was working in. So thank to new buyers,

half a year after that school, I quit my permanent position. It was really like,

for me, it was so, so exciting to,

to meet all these people and see how we can communicate. And I was,

we were discussing a lot about open science and sharing stuff,

and it was not the strategy of the group I was working in.

And that's why I quit.

And today I'm head of such a department and we are working in a different way.

We share things openly and it's possible. So that's, it's,

it's very exciting to see how these different approaches for running a core

facility, for running a group in bio image analysis work,

and as you just said it today, is very different ways it's possible to,

to get funding for these kind of opportunities for these kind of things.

And to me, it's yet unclear which is the right approach. Many approaches work,

many are financially possible. Um, but it's not so clear,

you cannot say the best way for running a co facilities like that. It's like,

it's just, we don't know it. They feel too young.

I, yeah, I, I think for microscopy corps there's good things, but for image,

for, for any type of analysis, actually, not just imaging analysis.

I'd say for genomic metabolomic, proteomic analysis, actually the,

the models are not well established yet. Yes. And, and I think,

think it probably fits your location, your funding potential.

But yeah, it's one, actually, I think it's one of the hardest areas to cost,

recover and to develop is in the,

i is in the analysis arena of developing that. Because

as an analogy, you are like lawyers, you know, your time has to be paid for,

you have to recover a salary on Coter, you are now freelance.

So this is blatantly obvious, but in the academic world,

it's more hard because we we're so used to collaborating,

just talking for free and helping. But that time costs.

And you know, if you are using, uh, if, I dunno,

if you're doing something physical and you pay someone,

someone to make a protein or spend hours on a, on, on an instrument,

it's clear what you're doing with them. In the case of image analysis,

it's far more amorphous. It's not, you don't see it in real time as much.

Uh, and it, and it's how long is a piece of string, you know, you,

it's hard to guess work out or estimate how long a piece of work is going to

take. So it does actually, there's

Actually, there's actually a risk also related to that. So what I think,

what I've done kind of wrong in the early days when I was working in this field,

um,

we basically only sent a script to the people who were requesting our support.

So today what they see is the final result,

and they see a 20 line script to a 20 line script.

So I think robot may have written it in an afternoon,

so it cannot have been so expensive.

That also leads later when it comes to who can actually,

who should be co-author on this, on this paper? I, yeah,

robot just wrote 20 lines of script that it's not such a big contribution,

right? But if I was sitting there for two weeks figuring out how to do this,

and then eventually wrote 20 lines of script, um, so the, the, the, the,

the actual work is not visible to the outside. There's a big risk here.

It's why nowadays what we, what we do more is like,

instead of providing a script to people and then hoping that they use it in the

right way, that's another topic. Um,

we write the script together with them so they learn something. Um,

and then also the next, the next project our collaborators want to do,

they could potentially do it without us because they acquire the skills.

So we are now more on the teaching and they were teaching the tools to the

people and showing, helping, helping them to help themselves.

And that allows us also to do more advanced projects in the future.

So this is the, basically the strategy here behind,

and then also gaining co-authorship by,

I was co-supervising this person in this data analysis task, um,

works much better,

much easier to justify because we understand each other better.

I understand better how the biological question has to be interpreted.

And the biologists, um,

I understands much better how the data analysis actually working and which are

there schools you can choose and which not.

I think, I think that's a really,

sometimes we take soundbites out of these podcasts,

and that's a really important one because I think that's a really good way to,

to make sure, sure, your efforts,

anyone's efforts is actually appreciated because you are, right.

I think that's the same again, if you're taking a microscope image,

I know as you've been doing the microscopy and you may do the microscopy for

someone and you give them back an image, which is seconds,

but the amount of optimization, troubleshooting,

designing the experiment to get to that point, which is a super simple image,

can be a long time. And, and I think people can appreciate that.

'cause people can get hands-on with the microscope.

And so they quite often pass it over 'cause it's beyond their skillsets if it

gets really complicated or ask for help, but you're right description,

they just pass it into, you know,

into yourselves and it becomes a black box and just comes back out. And they,

I think it's really good what you just said about involving them, showing them,

training them so they understand how hard it is. And Cota,

you do this freelance,

that how that's a brave step to step out to the world of

academia where you have a paid salary to go completely freelance. How,

how big a gamble was that?

I think I, I don't really recommend to no more people. I mean, uh,

so I I'm a bit weird person. I, I confess myself. I mean,

I'm very optimistic. So, uh,

so the first thing that to become a freelancing is, uh, I think, uh,

you have to be optimistic because you cannot expect salary at

all. So, uh, so I think, uh,

the point that you were mentioning about this, uh,

so how actually bio image analysts work with, uh,

biological academia, I mean, uh, so,

so the first point is that I think a bio image analysis is auto academia,

which is so, uh, research. So, uh,

it's very difficult to recognize, um,

from biology biologist that, um, doing some scripting,

uh, would be, uh, academic activity because,

so you may just making software, um, how can it be, uh, science,

something like this? So that's not true. And then, uh,

there's a different types of software, but, uh, um, uh,

so that, um, um, the software that you buy,

like Adobe Photoshop or Microsoft World, you can buy and use it.

And then, uh, so that, that is not,

so let's say that when you write, um, some paper,

you definitely need some editor. And then one of it is Microsoft Word.

You're using Microsoft Word,

but you don't ask Microsoft Word people.

So would you want to be a co-author? Right? But, uh, in that sense,

software is something that is really a tool like a pencil. Um,

so if you don't acknowledge the, the name of the pencil, uh,

maker in the paper,

because you just use that pencil for that writing the kind paper, right?

So, but in case of this scientific software, it matters more.

But of course there's a kind of, um, um, um, gradient of,

uh, um,

how important the specific name of involvement certain

software and author is in terms of scientific, uh, outcome.

But, um, in case of the scientific, uh, software,

I think it does matter. And then, uh, you cannot treat them like, uh,

Microsoft Word or other Adobe Photoshop because it matters

with the precision of your measurements. So, uh,

you write the name of the instrument, whichever you use for,

uh, scientific, uh, research, for example, that you,

you definitely could name the company name of the microscopes, right?

Size or Leica or Olympus doesn't exist, I don't know. But, uh,

um, um, Nyon or, uh, other companies, right?

So because, uh, it matters with the precision of your measurement.

And then likewise,

software that you use for image analysis has importance because of the precision

that you're doing and the algorithm that you using for treating numbers.

So, uh, um, in a similar way that, um,

when you are involved, uh,

in this research from the side of this image analysis,

there's a difference in the, um,

how you commit to that certain question so that if

you are committed just to provide, so, uh,

how about you can use this software just to measure your whatever.

So I think you don't have to be even acknowledged for, uh, doing that. But,

uh, if you are involved in writing something, yeah. Uh,

computer code even, um, not even, um, um, five lines. I,

I mean, uh, robots said 20 lines, but even five lines,

I think it matters with the whole project.

And then there's a much more deeper involvement, for example,

so you discuss with the person, biologist,

biome journalist, and bio B biologist talking, discussing about the question,

and then you start to say that, oh,

maybe we can measure this and have such a parameter there.

The biologist said, oh, that might be a good idea. Yeah.

So what I was often doing in Nimble was that talking like this,

and then you write on a piece of paper sometimes in the canteen on a piece of,

uh, napkins, paper napkins, and okay,

maybe you can put it like this and like that.

And then for whatever the treatment, it might go down in this part like this,

and then you can see the difference maybe statistically, and then you get,

get the results. So that is the involvement, right? Then you actually,

um, enlight the person to do some research.

And then there's an even deeper right now,

remodeling the experiment itself,

let's say that the person is taking a long time left B trying to figure out

difference in situation A and B, right? And then so,

uh, the person come up to you and say that, okay,

so we taking this TimeLapse and we want to analyze this TimeLapse maybe,

so that we want to see if there is a difference Yeah.

With this and that. Okay,

so I tell you then, then I would answer. But, so I,

looking at all those, uh, um, um, time lapse movies,

and it's not saying that I think you don't have to do time lapse movie,

you can just do two time points, the first one and the second one,

and then compare the situation like this in very precise manner.

And that would be give you a much better results. And I say, oh,

that's true. Uh, I don't have to determine lab naps microscope video, right?

And then, then you do the experiment,

and then you come back again and then do the analysis and get some results.

In this case, there's a kind of, you know, there kind of interaction that is,

uh, going, making something new, creating new value.

In that case, you're really informed, even if,

How, how do you, when you quote for this type of work,

Oh, okay, okay, you

Quote per day, per hour per day, or for a fixed job.

It depends. So that, uh, so I,

I stopped really working on, uh,

the real basic difficult research because then you, I cannot even quote,

but, uh, the, the problem might, so you, so you,

you have a data and then you see it, and, uh,

so it might take two days to just to want to get some results,

but eventually this can potentially become really longer.

It can take, yeah, like half a year or something because there's a new issue,

troubleshooting and so on. And then there's a small difference,

which actually happened to be the very important question.

So when you quote this, yeah, you cannot really quote, but, uh,

um, I, I actually made several, um, um,

mistakes in that way when I quoting, uh, so that, uh, um,

I quote when it becomes like an, um,

I mean,

I'm really sure that I can still do this in like a certain period of time,

then I quote, otherwise I say that, okay.

So that I cannot really expect the time, duration of the process, but, uh,

it might take infinite our money. Do you agree?

And if person agree that I would just start without quote

and then, uh, just say that, okay, so I work like, uh, um,

such a amount of time efforts, so that, could you pay me like this?

So that invoicing? Yeah.

And who, just, just very quickly, who are your main clients?

Is it academia or industry?

So it changed. So initially it was a lot of this, uh, academia,

but as I said, it becomes really ambiguous.

So I now tend to prefer to work with companies because, uh, of,

uh, it's much more clear with their goal.

Okay. Normally, yeah. Yeah. Which is why I envisaged probably to start with.

But I think it's fascinating that it can be done in academia. I think that's,

that's good. I have some quick, really quick fire questions.

So unmute your mics for those who have unmuted,

What, what is

It? So I'm gonna start with Elna. What was the first microscope you used?

It was, uh, SB five, like a state.

Okay. And Cota, what was your first microscope?

The first microscope? Yep. So, uh,

this is 1992.

I was using, um, Olympus.

I forgot, uh, it was already,

it was the first video microscope I used still this videotape,

which is custom made and then, uh, uh, setup.

But I, I, I really don't forgot the name of the, uh,

the microscope I used for this.

So it wasn't con it was just, that's just gonna be whitefield fluorescence,

potentially, yes. If that.

So probably an IX before the IX was about 1990.

Yeah, so, so I was taking, um, um,

micro tables, fixed samples.

This was already high tech. Yeah.

And Robert, what about yourself?

That's a tough question. I cannot recall the vendor, but certainly I got,

as a kid from my father, from my parents, I got a microscope for kid and I,

that was maybe at an age of 12 or 13 or something like that.

And a decade later I found it again and I tried to mount the camera on it,

and I think I failed it.

Oh, wait, does uh, keep the memories count too?

No, don't, don't, don't go there.

I can imagine what the toys around microscope were really like. No, no. It was

A real sound culture microscope when I checked the, um,

bacterial dishes under it, but it, I was, I think, sick. So,

Ah, six.

Yeah.

Oh my goodness.

My mother is in the diagnostics, uh, lab field, so

Yeah. Ah, hence the link through to it. Okay, next quick fire question.

I'm gonna start with Cota Image J on Aari.

Yeah. I don't know Nari yet, so, so well,

Well, the, these days, the, yes.

And you have two pe two of your fellow guests here have just said Image J. Are,

are you,

I mean, I was, I was using Image Day for, for,

for something like seven or eight years. Right.

And just to explain this a little bit, um, when I became Google Leader,

it was pretty clear that you cannot hire Java developers like Fiji plugin

developers on the job market, zero 10. Um,

but you can hire Python people and that's why we are a group of Python

developers who make NARI plugins,

Which is the key difference between Image J and the Parri is the,

the programming language for which and how long. I, I actually, I, uh,

met with the Janni ve recently,

and he's obviously very still image Jay. 'cause he's not,

although he's getting into Python a bit now.

How long do you think it'll be before we migrate everything from Msj

into Python and the par?

Everything. It's gonna happen.

It'll happen eventually.

It'll take forever to have everything. Like, just from a perspective,

do you

Not see something like chat G B T enabling that to happen?

Still people have to do it. Um,

I would say the majority of the tools which we love and Image J are already

available in NARI and the Python ecosystem. So you can transition,

um,

but it's still very different between these two platforms and potentially may

stay for a while. The installation process,

image J you download and it works with nari,

you have to struggle with Q environments and have to,

with dependencies and these kind of things. Um,

I'm afraid it'll stay like that for a while because there might not be an easy

technical solution to that. We need a social solution,

and the social solution might be better training this dependency problem,

and that software becomes more complicated. Um, over the years,

it's like unavoidable. The,

we are living in the 21st century and the computational revolution is

happening. It's not like that everything is done.

So it may become more complicated and the only way to deal with that is better

training.

Yeah. And I, I would, uh, for writing for scripting where you are,

yes, for the end user, you know,

which is someone like me who's not not into programming at all,

it's probably the interface and that being a very, uh,

single common, common type of interface, uh, which could really be,

have anything in the background as long as it's very user friendly,

which I guess is one of the big challenges is making those scripts into, well,

20 live scripts,

making them as user-friendly as possible and as least complex to,

to enable it to be widely uptaken. So how do you,

how do you publicize what you do?

How do you actually ensure that you've done that 20 line scripts, uh,

five line script quota? How do you make sure that your users, uh,

the wider users, not just the user done for,

but how does it get known by other people to be picked up?

How do you publicize it? Uh, who wants to pick that up first?

Maybe El ask, uh, come to yourself. How do you make sure, not just internally,

not just at Turkey,

but how do you make sure the wider community across the world across know that

it's available and what it's capable of?

Uh, the code that's, we write or we actually don't write code necessarily. I,

I mean, I'm coming from the core facility perspective,

so it's a bit different maybe than Robert and, and Kota, but, uh,

uh, I think it's, it's the training and awareness, um,

that makes a huge difference. Like if somebody doesn't know what's happening,

if it's a black box,

then of course they don't also know the effort that comes into it. Um,

but then if they know, uh, at least to some extent, um,

I think that that's the key point, turning point.

So get the core facilities to know about it and get them to then,

Uh, yes. And also users kind of

Won't be scoping. So I guess it's now down to you or me, uh,

in core facilities to make sure the users know it exists and

Exactly

Offer the solutions. But we need to know that that solution exists.

So Cota, that's another thing. Make sure I know it exists.

So, um, there's a different level of this, uh,

specificity and generality that, uh, um,

there's a general tools that can be used for many

different purposes,

and there's a very specific tools that is really focused on some specific

question, right? And then, uh, for those very, uh,

narrow bandwidth, uh, application, um,

it probably can use only once, uh, for that specific question.

And then not at all for everything. And then,

so there are some very general tools. For example, auto Two threshold, right?

Auto two threshold can be used in many different applications and then usable.

It's a generic in a sense. Yeah. So that, uh,

and then we shouldn't consider them all as a single type of tool.

It's rather that there's a different grade level scale of this, uh,

generality and specificity. But for those two, very general,

yeah, so we can advertise a lot,

put it everywhere as much as possible, and I do it. But, uh,

for those very specific application,

I think the use case in the future would be, uh,

almost zero, right? But it doesn't have much meaning to,

uh, advertise or tell other people that we did this.

It's rather that it's gonna be in the biological paper,

say that we use this script. Yeah. Actually it's

Good to,

it's that then it is the biological publication that then helps disseminate it

to, if there was anyone else,

it's going to be someone who's reading that paper that it becomes relevant to

Yes, that's,

Yeah. And then, so that was, uh, how I was initially trying to find out, uh,

how people doing image analysis. So you read Bal paper,

there's a method section, and if you're lucky enough, you can find a script,

right? And, uh, there was no GitHub before. Um,

so people are not used to publishing those, uh, pipelines. So, uh,

I mean,

so Bal paper was the media or this

speed, and then it still is.

Yeah, I think it still is.

And it's good points about GitHub giving access to it.

And we've got our a plus GitHub superstar. I,

I looked at your GitHub stats yesterday, Robert, which you, uh,

very high as an a plus GitHub contributor.

I dunno if that's the right term to use,

Whatever that might mean. Yes.

But I I I, anything else to add? We, we actually have four minutes left,

so anyone who's doing their exercising or ironing right now while listening to

the, to this, I've got, I've been,

it might take a bit more than four minutes because I've got a couple of killer

bits I want to bring in still. Uh, Robert,

is there anything you'd like to add for the determination? And yes,

We can come, I, I would like to come also back where the discussion started.

I think new bias and co buyers play a key role here in this context,

because many of us, in particular, the people who work in the core facilities,

we are developing similar things at the same time.

Like how many scripts have been written for counting nuclei?

I'm afraid it's billions of custom script for this particular purpose.

So by, by us meeting each other on new bias training schools,

on new Bias symposia, we actually learned that other people do similar things.

So we don't have to do it ourselves anymore,

we can just come back to the resources the others developed.

But that was like a key also for me personally, when I joined this community,

that was key for, for the success of new bias.

And that's why we also wrote something about that in the ZAYA proposal.

So one major thing we want to do, aside from founding the society, um,

is still bringing people together in these kind of workshops and potentially do

something like train a trainer.

So if I develop a new tool and then can train other trainers

in training people in using the tool, uh, the dissemination is like from,

from some perspective optimal. Um, so because you, you,

you share training materials,

that's what I do all day on Twitter and GitHub and whatnot. Um,

and in that way I can avoid that other people develop the same thing again.

Um,

so that is one of the major purposes of AYA that we can bring the trainer

together to exchange among each other to find common solutions for

pretty normal like problems which happen everywhere. Um, and then we have would,

we all would have time to develop new things and to develop new training

materials that would make our life easier.

So Robert, while I, whilst I have you,

'cause we have just got a few minutes left,

I have one picture that I managed to talk at research. Look, I I, I, I don't,

I don't stalk people with pictures, but I did have just one picture of you,

Robert.

Oh God, which

Is this one.

If I position myself correctly, it looks like, uh,

Martin's actually awarding you, me. Uh, but

so what do this, this award was any recent,

so what is the award that you are being given here?

Um, so that's the, the,

the data analysis and imaging award from the Royal Microscopy Society and what's

not in the picture. What also makes me very proud,

actually I got it together with Pete Bank.

So Pete got his award today before for a very similar reason. I got it. Um,

it's because we, we, we have like a lot of online materials where we,

where we train, where we, where we allow others to, to learn image analysis.

We both are very active on the image science form, image fp,

and we answer questions so we are acceptable and, uh,

help the community on a, on a global scale.

And I think that's what Pete and me got this price for also makes me really like

super proud to get together with Pete because we have such a very similar

mindset in this particular context.

I think that's a really important side. I,

so this just shows how far Cody you are back there, right?

Back in the, the early noughties. Robert, you started what,

mid noughties in this area? Elna into the tens, I guess,

into this environment? Yeah,

just how much it's developed in the recognition and importance of the work that

there are now, you know, prizes, uh,

recognition of the amount of work and how much that is needed and appreciated

by the core facilities. But actually now I'll take,

I did say I had a picture of you, COTA,

we go back in time to actually that conference you suggested.

Uh, so here, here's a picture of you Cota,

uh, in the rail museum in York. Uh,

I remember that exceptionally well because I, I,

so you've got Alison North actually. So podcast guest, uh,

in the early stages with Kurt Anderson, but also the Bena podcast, Chris Power,

uh, who you'll recognize by, oh, uh, through,

and of course Cota himself.

But Cota was also infamous at Elmi and actually Robert and

Elna. I bet you have no idea this, uh, on the banquet evenings

co would generally bring together the organizers.

So this is me and Joe Maron and would sing ah, and would ss

Sing,

Uh, to the whole, to the whole crowd, wherever the was.

So this is actually co singing, uh, with me and Joe, uh,

having no choice but to stand there next to him, listening to co to sing co.

What was the song? What, what did you actually sing

This? I don't remember, but I think so in this occasion,

I would say that was, uh, we'll meet that call.

It's a Japanese song. I think it was Japanese, right? Yes.

And, uh, so I was, uh, at this time, so among this,

uh, especially like, uh, Julian and other, these people,

um, we often singing this song. And then, uh, that was, uh, uh,

one of these, uh, yeah. So, uh, that was, uh,

yeah, I remember I was thinking, uh, and uh, this is a train museum,

right? Yes. Uh, yeah, coach. A lot of coach, old coach. And then, uh,

so I say this is a very good, uh, resonance to be there.

So I decided to sing,

Uh, certainly one of those unforgettable moments. Uh, as for the,

looking at what I was wearing, that just dates us right back. And actually I, I,

I didn't look that young down either, which is worrying really, before we go,

'cause we, we've gone past the hour.

I do need to just mention the importance of funding for new bias. New bias,

and whatever the new iteration will be.

And so who is funding that? At the moment?

We have,

we were very happy that we acquired funding from the CT Berg Initiative, um,

who is paying us now for the next two years. Um, so, uh,

we are lucky that we can hire a pertinent full time working on us,

with us as a community manager. Um,

and additionally we will have some funding for, for organizing these trainer,

trainer workshops I mentioned earlier, um, for getting people there,

in particular people from, from countries outside Europe.

So it doesn't make sense that we fund me visiting Heidelberg. That's like,

not necessary, but getting people from South America, from Africa, for example,

into our train the trainer workshops,

would allow us to spread the word about how we organize this community,

how we train, and then actually become a global,

become a global player.

Yeah, and you know what,

just thinking about Chan initiative gets mentioned so many times for different

initiatives,

but actually I think you are possibly one of the best benefactors,

not just from the funding of this, but maybe of the other things they're funded.

'cause by putting together and enabling bina, Biji, north America,

Latin America, BioGene Lab to the African initiatives,

you have all needs,

which means you now have point of contact to disseminate and to network in

otherwise actually getting pe to,

to raise awareness that you exist for people to come and join across

those other countries outside the elmi is a great place in.

So you've got that potential,

but actually for the other parts of the world that's not so easy. And c z i,

putting these networks together actually means it accelerates what you can do.

Even though the funding, even without funding, it just gives you a portal.

I actually, I've not really thought about that added benefit, but that

Actually, it's actually amazing how the Berg initiative,

how they also shape themselves and their services over the years,

because they are not just a funding agency which drops money at you and wants to

have a final report at some point. Now they, they really from time to time,

reach out and you have Zoom meetings with them and they ask you,

how can we help you so that you are successful?

And then through this discussions, there are these joint meetings, the, the,

with the other imaging communities, uh,

coming up and this is how we connected each other. So it's like clearly the,

the Chan Berg initiative who is really like making sure that we are successful

with the projects we propose. And there is, to my knowledge,

no other funding agency which work like that.

And so I I and thank you Chan Zuckerberg and all the other funders out there and

cost, you know, from the EU is, it's so vitally important for that early stage,

uh, of what didn't feel like an early stage, but he's now Harriett's growing.

You know, that was a seed for a lot of this and all the volunteers efforts,

you know,

all of you who put your time in at start quite often for free and finding added

time, trying to balance that with your work life balances and everything else.

Actually it, it's a community mm-hmm.

That puts in and probably puts in more contribution than any one thunder could

ever put in. Uh, so there's a big thanks there. One final question,

gotta ask, uh, starting with Elenas, what's the best conference?

That's new bias conferences, of course, there. So bias,

New bias,

Robert,

Definitely I would like to have this European white New Buyers Conference look

back like on an annual level.

At the moment we are becoming a bit like two national to regional.

From my perspective. I would like to have the big new bias meetings back. Yes.

So that's something we've gotta think about how to enable that and how,

where best ways of practice, guys, it has gone over the hour.

Actually, I've only asked half the questions I really wanted to find out.

So I've got a feeling we might be revisiting this topic very soon. Uh,

Robert Elna, COTA, thanks for everything you've done, uh,

for the community and everyone else involved with you bias and what they've done

for the community and where it's going in the future.

Thanks for everyone who's listened, uh, or watched. Uh,

it's worth watching if not just to see co to singing. Uh,

and please don't forget to subscribe and look,

listen back to all the other initiatives that are out there and people,

but please keep on doing what you're doing. It's awesome.

And thank you very much for joining us today. Thanks for having us. Thank you.

Thanks for support.

Thank you for listening to the microscopists,

a bite-sized bio podcast sponsored by Zeiss Microscopy.

To view all audio and video recordings from this series,

please visit bitesize bio.com/the-microscopists.