2021 Agronomix – Kidist Bioinformatics Interview
Earlier this year Kidist Kibret, PhD. sat down with Seed World Group President Shawn Brook for an interview about how Agronomix Software is tackling the important subject of bioinformatics. The video, which can be viewed below, outlines the plant breeding journey of Kidist and
Shawn Brook, President, Seed World: Kidist, thanks for sitting down with me today it’s great to see you.
Kidist Kibret, Breeding Software Professional, Agronomix Software Inc.: Thank you for inviting me to this interview.
SB: My pleasure Kidist, it’s great to see you I wish we could be sitting with each other, but we have a small issue of a global pandemic going on. So, I’m glad that we could sit down and chat via Zoom®, Kidist. Can you tell me a little bit about yourself and how did you first learn about Agronomix [Software]?
KK: I first learned about the company when I saw a job advertisement that suits my background. Actually i was recommended to apply by the job centre in Winnipeg and then i realized oh i used a very old
version of this software when i was a student. The previous AGROBASE and then then i learned more about the company after that. It was a company based on the background of programming and breeding and that’s a unique combination. I was like “Wow! This is unique”. I was very happy to join the company.
SB: Yeah, that’s awesome and good people there too
SB: Kidist, what’s your background? I know you’re a trained plant breeder but what’s your background?
KK: I started, my bachelor’s degree with plant science so that was an agriculture degree and focused on plant science. It included everything from agronomy to breeding in all different crops from fruits to coffee to grains. Everything was there. And then when I did my bachelor project [it] was a randomized trial with the carrot seed production. Then I met master students and phd students and they started explaining their projects to me. I was like “Oh well I want to do a master’s”. So, then I joined Addis Ababa University and did my applied genetics degree. Then afterwards I went to work on a fellowship in the UK and it turned into a masters by research. That was also crop science-based degree. And then I got a PhD opportunity in Germany.
I forgot to mention, before that also I was interested in computers as well, so I was like one foot in each. So, I did some computer courses when I was doing my bachelors, some programming, html some things like that. Then when I found the German project, I was very happy, because there was a large data [collection]. At that time, it was a huge thing and next generation sequencing was started and everybody was happy about that. But then there was a large amount of data. And nobody knew what to do with it. So I entered the field at that moment and I had to grow with the growing number of tools and [help in] choosing which tools to use. Which is optimized for these crops and so on?
That combined my interest in biology and also computing.
I did my postdoc in the University of Manitoba. When I came to Canada I came here as a skilled immigrant but i got the job before I came here so I knew the professors in the University of Manitoba. So, I joined them and then while I was working on my postdoc I met Dieter and then I decided “okay I’m stopping my postdoc this is exactly what suits my background this this job”. So, I came to Agronomix.
SB: Right. Perfect! That’s awesome well we’re glad you’re here.
Kidist, with a limit on our travel efforts these days, it’s not easy to stay truly engaged in what we’re doing, what’s something that currently gets you excited about the seed industry today?
KK: So, currently I’m following up on the new developments in bioinformatics and also with the new techniques of generating markers that’s really becoming very fine-tuned to the specific crop even to specific populations. That was the wish I had when I was doing my PhD. It’s not a generality, it should be specific and now that’s happening. I’m following it through webinars and with that webinar I discuss with people, experts in the field, so it’s very Interesting. It’s not as the same as going in person to a conference but still can be engaged. And then also I talk to a lot of breeders daily supporting them with Genovix®, with their breeding efforts. And that also is a great satisfaction. I get to learn different people’s breeding programs and have very nice discussions about the field. So, it’s very engaging right now even if it’s remote.
SB: Good for you. You said a word that until reasonably recently was a new one for me. And I continue to hear a growing conversation around bioinformatics. And the definition that I pulled, and I need you to tell me if I’ve got this right: bioinformatics is the is a discipline of biology and computer science concerned with the acquisition, storage, analysis and dissemination of biological data. Most often DNA and amino acid sequences. Why is this content so important for the plant breeding industry? And why is everybody talking about it, Kidist?
KK: So, you got the definition right. That’s exactly what it is. And then in plant breeding it’s more important because the breeding takes a long time. The conventional breeding takes time, and it takes space. You have to grow a lot of plants and [that’s] also expensive. So, when we develop markers that are switches, specific to that specific environment, to that crop, it’s slashing the labor cost and the time taken to develop new cultivars. And then even more precise, so precise breeding. It’s the basis, if I start from [the] background: The DNA is the basis everybody inherits, we narrate our characters, it’s influenced by environment. But it’s also partly our genetics. And then the amino acids are even more important in plant breeding than the genetics because plant genome is very complex. But then at the at the amino acid level we’re trying to capture the expressed genes.
That means, as in intermediate [level], that’s the usability of a lot of genes, they’re not being used in the plant genome. But in the ones that are really expressed in that environment so that especially for multi-environment marker development, like what’s expressed in specific for Canada for other places for a crop they say wheat, or this trait for that crop. So that’s very important and then [in] bioinformatics it’s very important to integrate our knowledge of genetics and genomics and transcriptomes even includes metabolic. You can, at the hormone level [get] all sort of biological information to make sense of how that expresses into yield or disease resistance. And then we can cut the time we take for the breeding. And then we also come up, maybe, even better quality plants done with conventional [breeding] but bioinformatics is very important because we have to make sense out of the data. [When] the sequence comes out it’s very large data. It’s a lot of: one genomic sequence, multiple times for accuracy. But then you need the algorithm behind to make sense of it. It’s like, for example, if it’s expressed how much is expressed. You just have to quantify everything for that gene; highly expressed in that specific environment is not expressed or lowly expressed amino acid for example. And all of that has to be precise and based on the biology of the plant. It has to make sense, for example: if it’s a colicloid plant it’s a different algorithm for the multiple copy of the genes. If it’s diploid, it’s different. If it’s human, it’s slightly different. So for each species we need to come up with some tools so that’s [where] computing is very important to get the precise information.
SB: To get it right yeah exactly. Okay and so Kidist, what value is Agronomix bringing to the discussion around bioinformatics and then what opportunities does that actually present?
KK: So, in here in Agronomix we have started a bioinformatics module and we already developed a genomic prediction tool and that’s doing prediction. You phenotype some plants in that population but then the rest you are predicting based on the genetic marker information [to see] how they will perform and that’s a very important tool once you have all your markers. It’s doing more than just a prediction so it will also free process if you have nucleotide format, it converts to numeric format and there’s also some parameters you can choose the threshold for (if there’s missing marker in the parent in the missing early genetic distortion) all those parameters so you can work which is more suitable with your model. You can model in your own population that will model for you. And the other advantage of using that is there is no need to export your breeding data and then use another tool. It takes a lot of time converting format from one to another bringing result back so that seamlessly happens in Genovix, so you have your marker and then you have your breeding information integrate that. Then you get your results inside Genovix and then you continue in there so no export-import. And then also the parallel computing we already have started with this module introduction. You can minimize, if it’s a large SNP (Single nucleotide polymorphism) dataset, you don’t have to wait for it. You [can] minimize or do other things in Genovix and get a pop up with your result. And then we have expanded Genovix to accommodate large genomic data, like terabytes! Some breeding companies might have a lot. Also, we are expanding to analyze large marker data or genomic data to mine the markers. So [you can see] we are we’re expanding it beyond that, so the breeders will have all rounded multi-disciplinary tools to advance their breeding.
Breeding Software BY Plant Breeders
SB: Excellent! So, the bioinformatics module is integrated directly into the Genovix software.
Right Okay excellent. That’s exciting. So, Agronomix has certainly gained a global reputation for breeding software, and it’s been built on a plant breeding software by plant breeders and why is that an important distinction to be made Kidist?
KK: Yeah, so being a programmer is one thing. You can become a really good programmer but to understand the needs of the breeder you also have to really understand, in depth, how the breeding works, how the plant biology, and the field design, all the statistics behind plant breeding. So, combining those two talents is it’s unique because those are two completely different disciplines that will take a lifetime to master each one. But when we try to combine these two in one company, and also even some of the some employees have both experience from computing and biology so that’s really a unique combination, and we can guide the programmers how to make it, every detail, every step because we know, we’ve been in the field, we’ve been in the lab. We know how competition, and everything is challenging when you do it on your own but then when we want to guide the programmers [to create] exactly what we want [then] they’re very happy. They can make something very useful. A very good example recently [was] we developed the Android, iOS and Windows mobile app. And that programmer who developed [it] we worked very closely with him. He showed me a feature and I suggested, for example I think back to when I’m in the field holding that mobile and I want to sometimes take all three traits at the same time and then some at some point I want to take this trait today and then the next day the other trait. So, he developed something to scroll by trait. And another option setting to scroll by plot. So, all these things where we are guiding them [to develop] exactly what the breeder faces and that’s what makes it very unique.
Bioinformatics: The Breeders’ “Magic Wand”
SB: Excellent! Good stuff. So, I’m gonna stretch your mind a little bit on this one Kidist. If you had a magic wand, what would be kind of one thing you would change as it relates to plant breeding for right now?
KK: Yeah, right now there is a lot of data for markers so it’s huge. It’s becoming very cheap to sequence crops, your populations. But really the big challenge is integrating that with your environmental data and fitting your crop genomics or that specific populations genetics because plants are very diverse. They are much more diverse than animals. So, you can’t use those tools for animal or human genetics this is it has to really be specific to plants. And then especially with expression with transcriptomes, that’s very useful but still a bit tricky and so in different environment if you can control the environment and see which marker is expressed. And then it can really be valuable in breeding so it’s very valuable. So, then the magic wand is bioinformatics that will mine so that’s the key and I’ve seen a lot of programs these days I have, but I still want more. More to be really like a magic wand where you can [use] your magic wand [to] just show you this marker is really good for selecting for this. Yeah, so we’re heading that way and that’s what we want to do with our bioinformatics project.
SB: Excellent, yeah excellent! That’s great thanks for giving me some insight into what’s going on over at Agronomix. Great to get to know you a little bit and thanks for taking the time to sit down with me.