I’m a systems neuroscientist, which means I look at networks in the brain and see how their activity relates to behaviour. This involves looking at the activity of lots of cells at once, so an important part of my research is developing new measuring techniques. Right now, I’m using one called CHIME. There’s a lot of data processing involved, so I knew that one day I might be able to apply these skills beyond my research.
Believe it or not, I was in the channel tunnel on my way back from Brussels. My colleague, Romeo Racz, and I were discussing how frustrating it is trying to find the scientific papers we need to do our research. We thought we might be able to find a solution using artificial intelligence (AI), which would have a deeper understanding of the search, rather than just using a few key words and entities.
We started to work on the idea as a weekend project and soon realised that if we wanted to pursue it, we would need support. That’s when we decided to bring the project into our work at the Crick, where we knew we’d find new collaborations and a ready-made group of potential users. With the help of the Translation team, we received an i2i (idea to innovation) grant, which we used to hire a programming specialist and get the concept off the ground.
Heron is a search engine that uses AI to extract the parts of research papers that scientists need to reproduce experiments or ask similar research questions, and conveniently compares them. It can tell scientists where to find the research that’s relevant to them, as well as relating it to their own data. It’s essentially “Google Maps for science” and we’ve built the interface so it really does look like a map – you can see the location of papers and how they relate to one another.
We’ve hit a wall with the huge volume of science out there. Currently, there are around 24 million published biomedical papers, with 2 million more published each year. Since only a small fraction are relevant to each scientist, it’s easy to miss important new research and data. With Heron, we want to overcome this challenge and make it easy to find the research we need.
The long-term goal is to change the way scientists communicate their research with one another. The current publication system is unsustainable and truthfully, I can’t see it lasting more than another 5-10 years. Heron aims to make scientific papers more visible, more interconnected and ultimately, move science forward.
When we first started, I knew very little about how to draw up a business plan, how to pitch to investors and all the legal obligations involved in running a company. It was incredibly useful to get this training in the workshops, talks and one-to-one sessions that KQ Labs offers. As well as being a very new, small company, everyone at Heron is a scientist, so getting input from people who have experience running a business was essential. Not only that, it was great talking to people who think differently to scientists and hearing what they thought about our idea.
However, for me, the most important part was actually getting ourselves out there and into the business world. The public-facing events hosted by KQ Labs gave us visibility and we met new partners that we’re still working with today.
My experiences over the past year have made me realise that my research is very important to me. I still want to go forward in academia, and hopefully I’ll be able to find an independent position in the near future.
But we’ve still got grand visions for Heron. We’ve just installed it onto the Crick’s systems and over the next year we’ll find out how useful it is in the daily work of our colleagues here. After that, I hope we can take it to institutes and universities around the world. Meanwhile, we’ll also be talking to industry about how we can provide services to them. I’m excited to see what happens!
