Early interest in chemistry and natural sciences brought Tõnis Laasfeld the main prize in the 2023 National Research Competition
Tõnis Laasfeld is a Research Fellow of Bioorganic Chemistry at the University of Tartu in the Institute of Chemistry and a data scientist at Proekspert. On August 29th of this year, under the supervision of professor of Bioorganic Chemistry Ago Rinken and Dr Leopold Parts, he successfully defended his doctoral thesis in Chemistry, titled "Integrating Image Analysis and Quantitative Modeling for a Holistic view of GPCR Ligand Binding Dynamics".
Tõnis describes himself as a bridge between Bioorganic Chemistry and Computer Science, as well as between academia and the private sector. During the interview, we asked Laasfeld how he ended up studying Chemistry at the University of Tartu, why he chose Bioorganic Chemistry, and what he thinks about winning the main prize of the 2023 National Research Competition.
Where does your interest in chemistry come from?
My first interest in and contact with chemistry probably arose when I was still in kindergarten, reading Jules Verne's "Mysterious Island" with my father. In the book, chemistry is used for survival on a deserted island. For example, soda was made from seaweed ash, and when combined with whale fat, both soap and glycerin were obtained through a saponification reaction. Glycerin was later turned into nitroglycerin, and the cave wall was blown up to create living quarters. I didn't quite understand the essence of chemistry at that time, but it seemed incredible and quite close to "magic". Of course, magic explained in sufficient detail is indistinguishable from science, and so the scientific interest in understanding various reactions developed later.
Why did you decide to study chemistry at the University of Tartu?
I had close contact with the professors and students at the University of Tartu and Chemicum during high school, thanks to various competitions. The people, labs, and atmosphere here were inspiring, and by the time of admission, there really wasn't any doubt in my mind. My high school chemistry teacher, Martin Saar, gave me a tremendous boost; his classes were exciting and fun. His teaching materials were of a high standard, and the practical work constantly showed how to apply the theory in practice.
What is your most memorable chemistry-related memory or story from your school days?
From school, I remember, for example, when we, together with a classmate who is now the Industry Liaison Officer of Estonia in CERN, Robert Aare, went to the chemistry lab during a lesson to "prepare for the chemistry competitions", but in reality, we spent half the time secretly burning magnesium and conducting other "totally scientific" experiments. Once, the teacher entered the lab precisely when a burning piece of magnesium landed on the laboratory table, burning a hole in it. It was probably the only time we got a proper scolding from Saar. We drew our conclusions, and from then on, someone always stood guard at the door.
Do you have colourful memories or stories stuck with you from studying at the Institute of Chemistry?
Of course, there are many. I remember the time of the Science Bus from my university days. We needed to design a new demonstration experiment for the show, and we decided to do acid-base titration with phenolphthalein as the indicator but with a "twist". Phenolphthalein is colourless in a highly alkaline environment, light pink in slightly alkaline, colourless in acidic, and orange in a very acidic environment. So, by dripping "solution A" into "solution B", one could create a colorless-pink-colorless-orange transition. Of course, "Solution B" was saturated sodium hydroxide, and "solution A" was concentrated sulfuric acid. In diluted form, such colour transitions don't work. Clearly, we were cautious, and nothing terrible happened. Still, under no circumstances should one try this experiment at home. Jürgen Vahter, our supervisor, liked the experiment, which surprised Maris-Johanna Tahk, as she suspected the experiment might have been "too basic". Considering the composition of the solutions, we were all very amused.
Interestingly, I haven't seen anyone else repeat this experiment before or after. This is a good example where adhering strictly to rules would have left a cool experiment undone. There were other incidents, such as the "opportunity" to go and fetch a forgotten giant oxygen tank from the bus during the performance in a school and then sprint to the third floor during the break to finish the show with a bang. I can't imagine what students not involved in the performance thought of this sight, but their eyes were wide open.
Where does your interest in Bioorganic chemistry come from? How did you end up in the department of bioorganic chemistry?
I am actually genuinely interested in all kinds of chemistry and other sciences as well. Part of the interest arose from the elective course "Biochemistry of Life" during high school. Later, I discovered that the authors of this course were members of the bioorganic chemistry department. I participated in chemistry, physics, mathematics, biology, and informatics competitions in high school. All subjects were interesting, and I wanted to continue with each in some way. Biochemistry already combines chemistry and biology. Listening to the presentations of all the labs in the chemistry building, it became clear that the bioorganic chemistry lab had all kinds of powerful microscopes under construction and in use, which needed a good understanding of physics and optics. As a bachelor's thesis, it was possible to work on a project where I could write software and apply mathematical modelling of the experimental results. So, in the bioorganic chemistry department, all interests could be put into operation as a unified whole, which suited me perfectly. In hindsight, the bachelor's thesis turned out to be strangely prophetic—I developed a method that allowed real-time monitoring of how a virus infects cells and spreads exponentially. If you pour a layer of agarose over the cells, the virus only spreads locally, and you get the classic plaque assay. The agarose layer worked the same way as social isolation during the coronavirus pandemic— it localized the infection but did not completely stop it. However, looking at the graphs of the spread of the coronavirus in early February 2020 and my bachelor's thesis graphs, it was clear that things would get out of hand very quickly and unexpectedly for most.
How did you come up with your doctoral thesis topic?
The topic developed over quite a long time. Partly exciting results and ideas emerged during my bachelor's and master's theses, which I couldn´t explore in depth. These were mainly related to the quantitative analysis of large volumes of microscopy images. On the other hand, during my master's studies, significant breakthroughs in machine learning and artificial neural networks occurred, which worked particularly well for image analysis and computer vision. At the same time, PerkinElmer (now Revvity) started a long-term collaboration with the University of Tartu, and we shared an interest in the same problem—how to detect and measure cells from bright-field microscope images. It became clear that the issues arising from the bioorganic chemistry lab and the solutions coming from computer science had to be combined. Also, Dr Max Keller's group at the University of Regensburg had synthesized a lot of exciting ligands for G-protein-coupled receptors, whose behaviour upon binding to the receptor did not follow the typical receptor theory. It was intriguing, and there was a strong desire to understand how the ligand binding to the receptor occurs, but this again needed substantial mathematical modelling. So, systems biology also had to be included in the topic. Finally, mathematical models were created based on measurements made from microscopy images, so all the pieces found their logical place.
What fascinated you about it?
There were many fascinating things. For example, when observing the result of an experiment, I often realized that I was "the first in the history of the entire universe", at least for a few minutes or hours, to hold this specific exciting new piece of knowledge. Even if scientific work starts to seem routine at some point, this thought still reminds me of how unique each scientist's work really is. Of course, creating new knowledge and technology to develop new and better drugs that could save millions of lives was also captivating. Although thousands of scientists participate in developing a single medicine and the preceding basic science, even if one were to distribute millions of saved lives between thousands of scientists, it is evident that working in pharmacology has a tremendous impact on many people's lives.
Do you remember your first fluorescence microscope image? What was on it?
Fortunately, the lab diary remembers almost everything. I took my first fluorescence microscopy image in the second year of my bachelor's studies of the Sf9 cell line labelled with a lovely orange-emitting membrane dye DiI. In the first image, the cells were practically invisible because the DiI solution taken from the freezer aggregates needs to be kept in an ultrasonic bath for some time before it can be successfully used. In fact, during this experiment, the idea arose that fluorescence dye may not always be necessary for detecting and measuring cells, and perhaps the same information can be obtained from bright-field images using machine learning algorithms. This way, time and materials could be saved. Later, this worked out, and several papers of my doctoral thesis use this technology.
How was your experience of writing your doctoral thesis? Was it a challenging process or an enjoyable one? What difficulties would you highlight?
Writing the doctoral dissertation took me around two months, with the last week at such a tempo that I managed to see the sunrise every day before going to sleep. But I didn't have to force myself; the thought just flowed and didn't stop until morning, so overall, it was quite enjoyable. Although the time was intense, it provided an excellent reason to set aside all other projects and distractions completely, and I could go very deep with specific ideas. The main problem I faced was balancing pharmacology, computer vision, and software development and interlinking these topics so that instead of a two or three-part thesis, a logical whole would be formed. Interestingly, this process led to the very first figure I created for the thesis ending up as the last figure in the final thesis, so certainly, it was not a straightforward process. In the end, I think linking the topics turned out quite well.
How did it feel to receive the main prize for your doctoral thesis?
The feeling was, of course, excellent, especially considering the tough competition. Wonderful works were submitted in the Sciences category, even from Harvard University. On the one hand, this shows that choosing the University of Tartu has justified itself, and on the other hand, it shows that if one systematically and passionately studies a topic, it is noticed. Indeed, this is a recognition for the entire labs of bioorganic chemistry and medical computer vision as well as the corresponding faculties and is an inspiration to continue working with the same topics. I also view this award as a sort of a guideline—apparently, something has been done quite correctly, and these skills and this kind of scientific method need to be passed on to the younger generation to ensure the continuity and further development of the fields I study.
What are you currently working on? How do you feel in your position? Do you think that your education at the Institute of Chemistry prepared you? Could you easily apply your education elsewhere?
Currently, I work as a Research Fellow at the chair of Bioorganic Chemistry at the Institute of Chemistry and continue researching G-protein-coupled receptors. At the same time, I am a data scientist at Proekspert, one of Estonia's first IT companies. At Proekspert, I mainly work on space technology in collaboration with the European Space Agency. Chemistry and related sciences are helpful and fundamentally necessary everywhere—for example, a satellite with a multispectral camera in space and a fluorescence microscope in the lab rely on the same photophysics. As a second example to calculate the optimal placement and number of solar panels on a roof using machine learning, it is good to know global energetics trends well, consider the development of electric cars and hydrogen technology, etc. So, in one way or another, all learned knowledge comes into play, whether it's biochemistry or some other branch of chemistry. The education at the Institute of Chemistry certainly prepared me well for data science too, and many data scientists have emerged from here—for example, Dr Karl Kaupmees, Dr Jaanus Liigand, Jaanus Burk, Ott Kekišev, and others. I think completing a PhD in chemistry is a good basis for any innovative initiative. Even if it's unrelated to chemistry, uncertainty and the unknown have become challenging instead of intimidating; instead of problems, you start seeing solutions, and any missing skills can be learned on the go.
How did you become a data scientist with a chemistry education?
I worked on software for several years while doing my thesis. At some point, I started reading more about data science and discovered that my activities were more in line with data science than just programming. During my PhD, I started dealing more systematically with neural networks in collaboration with Dr Dmytro Fishman, Dr Leopold Parts and PerkinElmer. At one point, there was a need for competence in computer vision at Proekspert, and I wanted to see how the IT sector works. Of course, such a change of topic was an excellent opportunity to validate some ideas that arose in the chemistry lab. After such a change, one starts to find connections between chemistry and data science at some generalization level. When a suitable level of generalization is reached, it also provides a direction for how detailed or general solutions to work on to so that these are widely applicable but do not become too abstract. "In the future, I will probably continue to be a "bridge" between biochemistry and computer science, as well as between academia and the private sector, facilitating technology transfer", said Tõnis.
These questions were posed to Tõnis Laasfeld by Maikki Moosus and Romet Peedumäe, marketing and communication specialists at the University of Tartu Chemistry Institute.
