Ola Spjuth wins Hjärnäpplet for robotic pharmaceutical lab

Ola Spjuth is a professor of pharmaceutical bioinformatics and has long worked with large data sets and calculations. A civil engineer by training, he specialises in molecular bioinformatics. His research career started with machine learning in collaboration with Astra Zeneca and other partners, but he soon realised that he needed to collect his own data.

“That’s when I got the opportunity to build my own lab here at the department. Actually, it all started with a small, automated microscope. Since I have an engineering background, I started looking at how we could robotise the experiments,” says Spjuth.

Automated microscope

No sooner said than done. The research team started working with cells that they exposed to different drugs and photographed. Initially, they followed the “cell painting” protocol developed by researchers at the Broad Institute in Boston. In short, it involves staining different parts of cells. When treated with medication, the cells are altered, and the staining makes it possible to see in detail how their organelles are affected. With automation, such experiments could be conducted on a large scale.

“We conduct 384 experiments simultaneously in each trial, and our lab has the capacity to do up to 20,000 experiments a week.”

As the lab was being built, developments in AI and deep learning were gaining momentum, enabling the research team to further refine the technology. Ola Spjuth started collaborating with Jordi Carreras Puigvert, a researcher from Karolinska Institutet who now works in the same research team.

“I came from machine learning and big data, while Jordi came from pharmaceutical development, microscopy and labs, so we’ve built this together.”

Rewarded with Hjärnäpplet

The results were so good that they also gave rise to a company: Phenaros Pharmaceuticals, located in Uppsala Science Park. And now Ola Spjuth is being awarded the Hjärnäpplet prize for his innovations.

“By creating AI-driven solutions and innovative companies, he shows how research-based technology development can lead to growth and social benefit”, says the committee’s justification.

“For me, a big part of our success comes down to building teams. I need engineers to build robotic labs, but also biologists and chemists to do experiments and drive questions. I need both expertise in pharmaceutical development and in software development, because we’ve built the control system for our robots and analysis tools ourselves.”

Today, around 18 people work in the research team. Two of them just joined the company, and at the same time Ola Spjuth is recruiting four new doctoral students for a major new research project to further develop the technology.

Potential for commercialisation

The journey from research to business took a few years, says Ola Spjuth.

“Construction of the lab started around 2017 or 2018. In 2020, I convinced Jordi to join my research team, and even then we started discussing whether there was a potential to commercialise this. We registered the company in the fall of 2021, but it wasn’t until 2023 that it really took off.”

What is your business concept?

“We’ve built a fully robotised lab that does a lot of experiments. From the beginning, I thought it was that large-scale capability that would be the most important. But I’ve since learned that the most important impact we get out of it is that we get data that is reproducible and of very high quality. Robots are great at doing exactly the same thing over and over again.”

Testing safety

They started with drug screening, but now focus entirely on drug safety evaluation. The company conducts large-scale experiments in which different types of cells are exposed to medications and collects data based on which AI models are trained.

In particular, they examine how new pharmaceutical products affect the liver. This is important, because almost all medications pass through the liver first, before being distributed in the body.

“Customers can send us substances with which we experiment in our robotic lab, and then we use our AI models to provide them with decision support. Our large-scale experiments provide very detailed information, which allows them to make better decisions. That’s our innovation,” says Ola Spjuth.

Industrial lab

The company's laboratory in Uppsala Science Park is built for industrial production and has ten times the capacity of the research lab at Uppsala Biomedical Centre. Last week, the company announced that it has acquired a biotech company in Cambridge and is changing its name to Pixl Bio.

“We now have ten employees in Uppsala and eight employees in Cambridge. Our new subsidiary has exclusive rights to produce relevant liver cell lines, which we will now combine with our technology for large-scale experiments and AI modelling. We also have an ISO-certified lab with high capacity to produce other advanced cell models, so this is a huge acceleration for us,” says Ola Spjuth.

Annica Hulth