Ghent start-up helps patients navigate oceans of medical data
While looking for specialist schooling for his son, Hans Constandt found search engines unreliable, so he created his own
Age of disqovery
“We envisioned having four big pharma or biotech companies as customers, and right now we have 12,” says founder and chief executive Hans Constandt. Earnings are also ahead of predictions. “Next year we will break even, I think, and then we will grow.”
Disqover is also picking up accolades. In November, it won first prize in the digital health category at a summit organised by the European Institute of Innovation and Technology, the EU’s flagship for new technology and entrepreneurship.
Big data is a fashionable area for new companies at present, but Constandt’s interest goes further back. He’s trained in medicine and biotechnology, but was also a bit of a computer geek, so he combined these interests by pursuing a career in bioinformatics.
After working at Ghent University, he moved over to the multinational pharmaceutical company Eli Lilly, where he worked on the information and technology demands of drug discovery.
Over time he saw that the challenge was not so much devising new ways of manipulating data, but connecting data created by different people – biologists, chemists, clinicians, toxicologists. “When you are making a drug, everybody needs to understand each other’s data, and it all has to work together,” he says.
But the spur to act on this insight was his young son, who appeared to have a learning disability. Constandt started to investigate what this might mean and looked into options for specialist schooling.
“It took me such a long time, and I’m a so-called expert!” he recalls. “So I took a sabbatical and started coding a platform myself to integrate all of that data.”
You need to know your keywords, and in pharma or complex areas where you may not be an expert, you don’t always know the keywords
The result was schoolKID, a non-profit platform that pulls together data on schools in Flanders and makes it easier to access. The response to this work was so positive that Constandt decided to apply the same approach to the more commercial world of drug development.
He set up Ontoforce in 2011, with backing from Flemish technology agency IWT and public sector research institute iMinds. The company’s first task was to convert the software built to sift school data into something more marketable.
“The engine I had made was built from open source components, and we said: We are going to make a limousine out of this. It has to be industry-ready, scalable and sustainable. So we started coding it from scratch.”
The key to the system is semantic searching, a way of sifting data that allows users to be specific about what they are looking for. The example Constandt uses is someone looking for a place to stay in France who types “Paris Hilton” into a search engine, only to get pages of celebrity gossip.
Adding “hotel” to the search gives a better result, but that’s only because the person knows what they want. “You need to know your keywords,” he says, “and in pharma or complex areas where you may not be an expert, you don’t always know the keywords.”
Free to use
For this to work well, data from different sources has to be connected up, so that searches can see through differences in terminology when it comes to diseases, drugs, genes and so on. Ontoforce did this for the most significant public databases of medical, chemical and biological research that it could find.
“We took all of that data – about eight billion entities – and we merged it together into a very scalable and easy-to-use platform,” Constandt explains. Since the data is connected, users can also make connections, moving from clinical studies, for example, to the underlying biology or to the chemistry of the drugs involved. Information can also be found about the researchers, such as their publication records and other interests.
We took all of that data, about eight billion entities, and we merged it together into a very scalable and easy-to-use platform
Since October, Disqover has been free to use under a community licence, with around 1,100 people currently signed up. Where Ontoforce makes money is adding functionality – for instance allowing people to save and export all their searches – and by licensing the system to pharmaceutical companies.
Under this arrangement, commercial databases can be added to the system alongside internal data, allowing employees to factor in company research and resources.
Ontoforce now has a staff of 14, mixing people with entrepreneurial expertise with those skilled in IT and bioinformatics. In June last year it raised an additional €2.1 million from the Life Science Innovation Fund in the US and its existing investors, such as iMinds and Flemish investment companies Sofi and LRM.
There is a strong interest from the US, but Constandt will keep the company’s research and development in Flanders. “We are very happy here. The talent is here, and we can find the money we need. It’s only sales that are going much faster in the US.”