Dan Greenfield on how he came to co-found a successful genomics company and why Cambridge was the ideal place to do it.
I would not have been able to start my company if I had not been in Cambridge. I cannot imagine starting this company in any other place.Dan Greenfield
When he finished his PhD Dan Greenfield  wanted to keep one foot in academia and the other in business. In the end, due to UK visa restrictions, he was forced to choose. He opted for business. That choice has proven fortuitous as he has co-founded a successful genomics company which aims to bring down the costs of genomics collaboration and bring its benefits to as many people as possible.
Dan became interested in the rapid growth in genomics and the potential it has for personalised medicine while he was at Cambridge University. He noticed that, while the cost of sequencing was becoming more economical, that of data transfer and storage was still high. “We were keen to explore how to address this and help democratise genomics research,” he says.
Storleap, the company he and a fellow computer scientist founded in late 2010 aimed to develop advanced, next-generation storage and synchronisation technologies to existing IT infrastructure, but the process of applying it to genomics and commercialising was more of a challenge. In 2015 Dan started PetaGene to do this and by 2016 he had exhibited at Bio-IT World in the US and won Best of Show. That brought lots of interest and investment, enabling the company to scale up and properly commercialise. It has since managed to win major customers, including AstraZeneca and is starting to work with national health programmes and systems. PetaGene now has 17 members of staff and offices in the US and the UK as well as individuals working in Serbia, India and Spain. Some members of the team and advisers have been Gates Cambridge Scholars, like Dan. “I am very proud of everything we have achieved,” he states.
“We work to reduce the costs of genomics, but also help organisations with data security so they can protect their confidential genomic data. A lot of organisations have data, but cannot share it because they cannot ensure its confidentiality and because it is too bulky. Our products make research collaboration easier and help to democratise the field.”
From an early age Dan has had an entrepreneur’s interest in problem solving. Born in Sydney, Australia, he found it hard to rote-learn things as a child so he developed his own ways to solve problems instead. That approach got him accepted to study Computer Engineering at the University of New South Wales.
Dan had started computer programming when he was just seven after his father bought him an old Apple computer. “I spent a lot of time developing things. I didn’t play games so much as write them,” he says. At secondary school he was one of a team of three who represented Australia in an international programming competition in Pakistan. He also did the Australian Physics Olympiad.
The University of New South Wales allowed him to skip first-year Physics modules because of his advanced skills. He knew at the time that he wanted to do science and “invent and fix things and solve grand problems” and says he was inspired by entrepreneurs like Bill Gates who “had built businesses that improved the world and then put their money into charity”. “I wanted to be both an academic and an entrepreneur,” he says.
As part of his degree, Dan had to do work experience. Usually this was done locally, but there were not many hi-tech firms locally so Dan chose instead to do an exchange with the University of Illinois Urbana-Champaign after an internship offer from Microsoft. Through attending a careers fair while he was in the US he interviewed for the company Silicon Graphics and spent another year working in Silicon Valley for them, initially as an intern. He ended up, however, with his own office, working on their next generation graphics processor, designing a new capability that is now essential in machine learning.
When he returned to Australia, Dan continued to work for the company 20 hours a week while still finishing his degree full time. After graduating he moved to Silicon Valley and then Hawaii to work for a start-up, but after this was acquired by nVidia he decided that he wanted to do a PhD.
Dan applied to Cambridge, but just missed the deadline after turning down a tempting offer of a managerial position in Silicon Valley. So instead he started a PhD in Bioinformatics while he was waiting on his Cambridge application, knowing he could downgrade to a master’s if necessary. He managed to finish his two-year master’s in a year while also continuing to do his consultancy work. At Cambridge he was keen to continue the work he had begun for his undergraduate dissertation on connecting processors on a microchip to larger networks.
He started working on networking, but his interest in ideas soon took him down another path, looking at the mathematics of how the human brain is organised. “Everything excites me – I like to bring disparate ideas together,” he says.
His PhD focused on the physical locality of computational networks and looked at how theory can explain some of the mysteries of mammalian neuronal networks and perhaps help to explain other natural phenomena where physical position matters such as social, epidemic, financial and traffic networks. This work was awarded the BCS/CHPC Prize for the top computer science dissertation in the UK.
Dan admits that, like many scholars, he felt a bit of an imposter when he went to the Gates Cambridge orientation. Although he had helped to organise a science camp for hearing impaired children at school and university and worked with Engineers Without Borders during his master’s, he met people who had founded their own NGO. “I thought ‘these people are next level’,” he laughs. He grew to love the Gates Cambridge community and says it and Cambridge have inspired him. He credits Cambridge – both the people he met and Cambridge’s history as the birthplace and an international centre of genomics – with giving him the idea of applying computer science to the human brain.
He adds: “I would not have been able to start this company if I had not been in Cambridge. It gave me access to the people who gave me the idea for the company, exposed me to some of the problems we aim to solve and meant I was in the right place to hire the right people to work with us. I cannot imagine starting this company in any other place.”
- 2005 PhD Computer Science
- Trinity Hall
(Update: I am now CEO of PetaGene. We tackle challenges in Personalised Medicine, making unwieldy genomic data from sequencers smaller, better and faster, to reduce costs, improve analysis and speed up collaboration.) My PhD research developed models for the physical locality of networks. Locality is fundamentally important for the performance of future computer systems with thousands of processors on a chip, but not much is fundamentally known about it. What is very exciting is that in collaborations with the Brain Mapping Institute, we've also found the theory can explain some mysteries of mammalian neuronal networks and we believe it may help to explain other natural phenomena where physical position matters such as social, epidemic, financial, and traffic networks.