I first came across BenchSci, previously Scinapsis, as a contributor where I wrote a few articles on life science professionals and STEM careers. The concept behind BenchSci seems obvious, a concept that one would expect to have been implemented in some way. Based on an initial assessment of BenchSci I have provided a “Scientific Method” approach to highlight key findings:
|To build a search engine capable of housing all commercial and published proteins for research.||Allow researchers to easily navigate and locate their proteins of interest through a search engine, and provide all the relevant information, saving time and eliminating hours of literature review.||Machine learning algorithms, data mining.||
||Expansion to other products, market expansion, knowledge translation and open data access.|
To put it simply, BenchSci is a search engine that offers its users the opportunity to find their desired antibodies (proteins that play an integral role in the immune system) without having to rely on an extensive literature review or paper trails. To give you some context, as a researcher you might be interested in studying a specific protein within a specific cell line, but you’d first like to know which antibody will give you the best results. Normally, you would read through literature and visit Pubmed scrolling through the published material which may or may not contain experiments that used the antibody you are interested in. Sometimes, these papers will publish images of their results (ex; a blot from protein analysis) .
Based on the protein the paper has used, you then turn to vendors who sell the products or reagents of interest. From this point, there is quite a bit of trial and error, your protein may work in your desired experiments, it may work after optimization or it may not work at all. To speed up the process, BenchSci aims to eliminate the article search and has created a search engine where you are prompted to type in your protein of interest. The search engine allows you customize and tailor your search (ex; specific technique you will use the protein with, cell lines, tissue), where just like Google you are returned with hits that best match your search, particularly articles in which your protein of interest has been studied in. “There definitely is an advantage when you have a search engine that allows its users to customize their search and retrieve highly specific hits” Maurice explained. “Researchers and scientists are looking for specific reagents and products, so having that specificity is important.”
“But how do you convince investors of your idea? How do you explain all this scientific terminology to investors?” I asked. Here, Maurice caught the meaning of what I meant. “It’s not easy. It definitely took some time to be able to explain what we were doing and it also helped once we became better at communicating our service to different audiences. That came with practice.” I wasn’t surprised, being able to take largely scientific services or concepts with all the technical jargon and terminology to investors is no easy task. How best would you be able to relay to an investor that your service will cut down the time needed to find a protein of interest that will guarantee better results when running a western blot or a qPCR? Knowledge translation, the ability to communicate the relevance or the potential of an idea into a product is an important field, gaining increased momentum within the start-up community, particularly the scientific community.
Which brought me to my next line of thought, how would BenchSci promise return on investment and how would they be able to attract future investors. “Expansion” noted Maurice. “If we can have this service for antibodies, we can also use it for other reagents and other products that the research community requires, if we are going to stay relevant, we will need to expand.” And expand they’ll have to. For now, they do have the advantage of incorporating images in their search engine, but moving ahead it won’t be enough to cater only to scientists and labs.
Two things impressed me particularly with BenchSci:
1) being that their core team was made up of a loyal and committed group of PhD graduates and students who found a problem in their everyday settings and set out to solve it. “It was a great opportunity for us to be involved in and run something that belonged to us” mentioned Maurice. “I think being an entrepreneur is very much related to being a PhD student, there’s a lot of uncertainty, risk, problem-solving, a huge amount of self-management and a natural desire to make things better.”
And 2), machine learning. The entire system currently relies on an original and highly specific supervised set of algorithms that the team spent the last few years editing and refining. The algorithms are capable of identifying specific terms related to literature review searches, starting with an input data the algorithm is trained to yield the required output or the search results.
This brings me back to a start-up competition event that I recently attended in Toronto, where a venture capitalist mentioned (in their opinion) that the most important factors that would hold any early start up together are: the team, the leader and the ability to define a market. I wondered about whether BenchSci had determined these factors. “I think BenchSci’s strengths are in three main points” noted Maurice. “Our strong and devoted team, our leadership and our algorithms which are highly selective.”
In a way, the selective nature of the algorithms determine the niche market that BenchSci currently targets, which are researchers and the academic community, and for now Toronto holds that market. It is an unprecedented advantage, what with the recent attention and focus on machine learning in health technology, the multitude of research labs, and the launch of the Vector Institute in Toronto. But perhaps more importantly, amidst all the innovation and hard work being put out there, longevity could be the true currency. Being able to stay relevant within an ever changing market could be the ultimate determinate as to whether BenchSci and other scientific tech start-ups will serve their users and supporters in the long run.