Following the data

Mikael Huss is one of the researchers who set up the bioinformatics part of the genomics platform of SciLifeLab. Now, he will coordinate a new initiative at SciLifeLab to combine data from different kind of experiments for the same model system.


In the mid nineties, when Mikael Huss started studying molecular biotechnology at Uppsala University, bioinformatics as a field did not really exist, but Mikael was already then interested in combining technology and biology. He did a lot of programming in his spare time and was especially fascinated by neural networks and genetic algorithms, which is a way to optimize things inspired by the evolution. At the time, he also took a course about pattern recognition, which was a real eye-opener for him.

“Pattern recognition is really statistics but in a more fancy package. It made me interested in big data and how you can apply pattern recognition in biology in different ways.” Says Mikael Huss.

At his blog Follow the data Mikael today writes about computer analyses of large data sets, often with biological applications. Recently, a lot of the posts have been covering a new hot subject called deep learning and how it can be used in biology. In deep learning you use several layers of networks, which together make a summary of data. For example, Google let a computer read billions of pictures on the Internet and extracted the most common pattern, which turned out to be a cats face. The principle mimics the human visual system where different cells in the brain recognize different elements in what we see.

Learned about NGS by chance

Mikael has always had trouble limiting himself to one area, may it be hobbies, studies or his job. For example, he took a long break in his biotechnology studies to study Chinese in China. He has also vented his passion for languages in writing the bioinformatics section in the business-to-business magazine Life Science Sweden on a regular basis for several years.

“In school, I liked to learn about cell functions but I did not enjoy being in the lab and I don’t think I was very good at it either. I am quite impatient and do not like repetitive work, I get bored too easily. At SciLifeLab it is pretty good since I work in a lot of projects at the same time and we are also quite free in what we do.”

After his dissertation at KTH Royal Institute of Technology in 2007 Mikael did his postdoc at Genome Institute of Singapore. At the institute they already used next generation sequencing at that time which has also become one of the biggest things a SciLifeLab.

“It was a pure coincidence. My wife and me chose to move to Singapore since we both speak Chinese. I didn’t even know NGS existed and I came early directly in the frontline of it. My colleagues and I fumbled a bit in the dark, trying to understand how to work with the generated data with bioinformatics.”

When they moved back to Sweden a few years later, Mikael had heard that SciLifeLab was about to start and contacted the center to ask for employment. Together with two other bioinformaticians he set up the bioinformatics part of the genomics platform in 2010.

Knowledge transfer important

Since a couple of years Mikael now works in the bioinformatics platform WABI at SciLifeLab where he provides bioinformatics support to research groups for a longer time. Right now he is involved in one project coordinated by researchers here at SciLifeLab, one at Karolinska Institutet and one at Umeå University.

“We often have two bioinformaticians from our side and ask that there is one person in the research group who can capture the knowledge we have, and who can continue on their own when the time we have in the project is over.”

Right now, Mikael is preparing a new “big data” initiative in WABI, which he will coordinate. Within this initiative that focuses on integrative bioinformatics, WABI will hire 4 persons; 2 at Chalmers and 2 in Stockholm.

“At SciLifeLab we have a lot of competence and techniques to generate high throughput data like DNA-sequencing, mass spectrometry and imaging. But a reoccurring complaint from the researchers is that not enough is done to combine data from these different levels. Therefore a common approach is needed to see what we can find if we put DNA, RNA and proteomics data together with metabolomics measures of the same system or cell line. This involves bioinformatics method development to create the necessary tools to combine data. The plan is also that there will be a special track for integrative projects when you apply for WABI support where you get more hours to ensure that you can do a sincere attempt to put data together.”

What you did not know about Mikael: He trains his son’s soccer team and likes to play chess. He also likes to watch “weird”, independent experimental film, preferably from Asia.