Face to Face with Thomas Leung, CSO, BenchSci.
The first day I started my postdoc in the Bremner lab, I remember talking to Tom, a graduate student working in the “Epigenetics” wing of the lab. Being fun-loving and most importantly coffee loving, we instantly bonded and formed a team…doing Science and talking non-sense. I witnessed the BenchSci growth closely…it is amazing how Tom took his idea forward, pursued relentlessly and now successfully launched his startup, raising money from both angel investors and VCs. Within a short span, the BenchSci team won University of Toronto Banting and Best Centre for Innovation and Entrepreneurship (BBCIE) Fellowship 2016, Ontario Centre of Excellence (OCE) Smart Seed 2016 and the Brightlane Entrepreneurship Award (BEA) 2016. To inspire potential start up seekers in CSG, I interviewed Tom recently about his journey with BenchSci.
ME: Tell me about BenchSci?
TL: BenchSci is a machine learning software that analyzes and decodes scientific papers to extract antibody usage data in the form of figures. These figures are then further indexed and aggregated to make them easily accessible to the research community.
ME: So, how did the idea get started?
TL: During my PhD, one day I was planning a new experiment, which required a lot of new antibodies for this huge Western blot. I was sitting in front of my computer, using conventional search engines and looking through PubMed to search for antibodies that have been validated in peer reviewed papers. After many hours, I thought to myself, “wouldn’t it be nice if there was a database somewhere that I can just input my favorite protein and I will be able to see all papers produced with different commercial antibodies against that protein?”. I started looking online and realized that such database does not exist, so I decided to build one on my own.
ME: How did you go about it? What’s the process involved and how did you form a team?
TL: To build this massive database, I know that I am going to need someone with superb programming expertise. My whole academic career was in Life Science and I do not know many people in Computer Science. I know that UofT is a great place with awesome ComSci talents, so I logged into my LinkedIn account and typed in “UofT, programming”. The first result was David Chen, who became our Chief Technology Officer. Amazingly, David is both an adept programmer and a PhD researcher in Neuroscience. I invited him out for a drink and we chatted for many hours and that’s how the team got started. I continue to look around UofT and assembled an awesome team right here, including our Chief Executive Officer, Liran Belenzon, MBA from Rotman, our Chief Database Officer, Elvis Wianda, PhD from Medical Biophysics, and our Community Architect, Maurice Shen, PhD from Pharmacology.
ME: How did your exposure in University of Toronto (U of T) help you in this pursuit, from a lab to a startup?
TL: UofT has many incubators aiming to nurture and support new ideas. We took advantage of this great opportunity and went to a few of these wonderful incubators such as the Hatchery at the Department of Engineering, the Creative Destruction Lab at Rotman Business School, and H2i at the Faculty of Medicine. Also, being a research scientist myself means I was able to talk to many professors and researchers to get valuable feedbacks and comments. For instance, our scientific advisors Dr. Jim Woodgett, Director of Research at Mount Sinai Hospital, Dr. Ruth Ross, Chairwoman of the Pharmacology Department and Dr. Ioannis Prassas, Staff Scientists at Mount Sinai, gave us many great suggestions that helped us develop and improve the BenchSci platform to better serve the scientific community.
ME: How is your platform different from the several antibody validation websites that already exist, like Antibodyreview, Biocompare, etc…?
TL: We are a true validation platform, meaning that we directly showcase experimental usage validation of antibodies in peer-reviewed journals. As a researcher, I realize the importance of seeing a figure more than anything, that’s why BenchSci is designed to show scientists antibody evidence-of-use directly in the form of figures, and not just a mere citation number.
ME: So right now what’s the design of your BenchSci website?
TL: It is very straightforward. All you need to do is go to our platform, type in the protein you are interested in and press, “enter”. We will show you a list of figures produced by commercial antibodies that target your protein of interest. You can continue to narrow down your search to fit your experimental criteria by applying multiple layers of filter including technique, tissues, cell lines, and disease models. We have a demo video on our website at www.BenchSci.com. For one minute of your time you will immediately realize how simple it is to use BenchSci.
ME: Do you plan to move to other reagents, other than antibodies?
TL: Yes for sure. BenchSci is a powerful software that can decode scientific papers, and we are also planning to target other experimental reagents that require validation information before making a purchasing decision.
ME: Is the software the end product that you will sell, if so, what is your future direction after that?
TL: We are offering BenchSci free to use for all research scientists. We truly believe that BenchSci would be helpful for researchers around the world. Many PhD students that we talked to had one recurring comment: “oh how I wish I have something like this earlier in my career!”.
ME: You mentioned to me that you are the CSO, but not the CEO of the company, although the idea is yours. For scientists like me who do know much about startups, can you describe how these titles work out? What or who decides these things?
TL: Each of our founders plays very specific role in the company. As the Scientific Officer, I am responsible for all things life science and biology related during product development, from backend data collection logic to frontend user interface search mechanics. I am not directly involved in the coding (which is done by David and Elvis), I design the scientific reasoning behind the code. Our CEO Liran is responsible for all things on the business side of our company. It is a triangle: Science, Technology and Business, each of the founder’s specialties in each of these components makes the team strong.
ME: There are two types of people who are getting into startups. One kind, like you, start with your own idea. On the other hand, I was attending a talk recently and the guy wanted to start a startup and he did some research on what’s hot right now and came up with an idea and went about it. According to you, which type is more sustainable? Or do you think both will work the same?
TL: The story that you build from the idea is the important element. A good story will resonate with people and bring more impact to the idea. However, the idea can either be something that took place in your dream, or something that was triggered after hearing another person’s seminar. The only difference is that, if you are creating a solution to a problem close to yourself, it is easier to convince others the value of your solution. It is more credible for a cell biology scientist to create a solution for the reagent problem than, for instance, an outsider from mechanical engineering. Let’s say I realized this terrible traffic problem on the highway and wanted to build a transportation system to solve this problem. This idea itself might be very good, but since I have no computer or engineering background, it would be more difficult for me to convince people about this idea.
ME: Finally, do you have any advice for beginners, who want to start a startup?
TL: Imagine a road parked full of cars, looking for a parking space is not going to be possible. If your car is your idea and the road is the market, with so many other solutions already out there, it would be tough for your idea and product to develop and grow. To build a startup, good “product-market” fit is important. Do not try to find parking space on a road already filled with cars. Instead, create solutions for problems that do not yet have a good solution. Maurice, our Community Architect, wrote a very good article for students who are thinking about startup, you can read more here.
You can find more information about BenchSci, see the following:
Company info: http://www.benchsci.com/about/.
Demo/Introduction video on www.benchsci.com
Company statement: http://blog.benchsci.com/2016/09/15/the-benchsci-story/ “The BenchSci Story”
http://www.theglobeandmail.com/report-on-business/small-business/science-startups-make-research-faster-cheaper-more-accurate/article32270645/ “Science startups make research faster, cheaper, more accurate”
About Thomas Leung:
Tom Leung completed his MSc in Virology and PhD in Epigenetics at the University of Toronto. For his PhD thesis, he investigated the molecular mechanism of repressive genetic bookmarking during cellular division and the potential application of reversing these bookmarks as alternative cancer therapeutic approaches. As a molecular biology research scientist, Tom experienced first hand the inefficient organization of biomedical publications.
Tom is very passionate about the development of a solution to better organize the vast amount of data in scientific literature in order to bring the most relevant information to scientists to facilitate the next big biomedical breakthrough.
About Manoja Eswara:
Manoja did her PhD from University of Guelph, Canada, where she worked on unraveling nuclear cytoplasmic transport pathways for transfer RNAs (tRNAs). Currently, she is doing Postdoctoral fellowship at LTRI, Canada, on Cancer Molecular biology and Epigenetics. Her work is focused on understanding the epigenetic factors involved in regulating replication and gene expression in Cancer cells and the potential use of small molecule inhibitors targeting them as Cancer therapeutics.
Featured image source: Pixabay
This work by ClubSciWri is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.