If you’re a research scientist, you’ve probably experienced a failed experiment. A lot of times, this is caused by poor antibody selection. On top of the feeling of defeat and frustrations of troubleshooting, wrong antibodies waste time, money, and precious samples. If only there was a way to choose the right antibody for your experiments the first time around.
The BenchSci office space
Enter BenchSci. BenchSci is a Toronto-based start-up, founded in 2015, has developing a highly efficient antibody search platform. BenchSci uses artificial intelligence and machine learning to decode scientific papers and collect antibody data that are unbiased and experimental context-specific. The search platform is algorithm-driven and researchers can search for the antibody to their protein of interest and receive a comprehensive list of commercial antibodies. These searches are accompanied with corresponding figures from scientific literature, and even link to antibody purchasing via specified vendors. Importantly, the platform has numerous filters to let researchers narrow down the search results, identify the antibody that best fits their experimental context and successfully find the best antibody for their assays. Oh, and the platform is free for academic users!
Nucleus: A meeting room in the BenchSci office
We had the opportunity to visit BenchSci’s headquarters in Toronto and speak to a few of the talented individuals driving this amazing work. BenchSci is a vibrant, open-concept space, facilitating collaborations between teams. Upon arriving, Maurice Shen, Head of Academic Relations, welcomed us and guided us into a meeting room called “Golgi” (all the meeting rooms are named after organelles!). He then offered us coffee served in beakers, joking about how, at BenchSci, their caffeine consumption is easily measurable.
Maurice completed his PhD at the University of Toronto in Pharmacology and Toxicology before becoming one of the first individuals to join BenchSci’s team. “Our platform was designed to understand which antibodies are available and how they’re being used in the literature. We then make these data available on a single platform for scientists to find the most relevant antibodies for their studies,” Maurice explained.
While this concept seems intuitive enough, it took over two years to develop the artificial intelligence algorithm to make this platform possible. Casandra Mangroo has been involved in the behind-the-scenes work since not long after the company’s inception. Casandra completed her PhD at the University of Toronto in the department of Laboratory Medicine and Pathobiology and has transitioned from a Scientific Associate at BenchSci to the current Head of Science. She explained some of the early challenges they faced: “A lot of research is closed access, so we had to work with publishers to gain access to research articles. There’s a lot of ambiguity and complexity in research where genes and proteins can have multiple aliases, so we had to merge that data. Finally, vendors don’t have a standardized labeling system for their products; if the target species of an antibody is human, this can be labeled ‘h’ for some vendors and ‘hu’ for others.” BenchSci initially trained a team of PhD scientists to curate, label and consolidate these data, which ultimately trained the algorithm to do these operations on its own. “Over time, we automated as much of the process as we could.” Casandra explained.
Maurice Shen (Head of Academic Relations), Casandra Mangroo (Head of Science), and Matan Berson (Head of User Experience)
Now, Casandra oversees the science team to ensure the integrity and comprehensiveness of data on the platform by improving the algorithm, updating information and ensuring the knowledge graph is always growing. She also travels with the sales team to demonstrate the functionality of the platform to show how it can add value to scientists’ research. She added, “Seeing scientists use our platform face-to-face motivates me everyday, because I know the work going on behind the scenes is making an impact in so many different fields.”
While gathering and processing huge amount of antibody data was BenchSci’s first hurdle, designing a user-friendly platform for scientists was another challenge. Matan Berson, Head of User Experience, was an integral part of this process. Another University of Toronto graduate, Matan completed his MSc in Biomedical Communications before joining BenchSci. “With all these data come challenges; the biggest being meeting the specific needs of scientists while providing the value that we’re promising,” he shared. Matan’s objective was to understand the user experience, which varies greatly due to a wide demographic of scientists, and display the information that is most relevant for antibody selection on the search platform. The ongoing goal is to ensure BenchSci’s antibody search engine is intuitive to all of its users. Because the platform is being used by new research students, seasoned academic investigators and industry scientists, this is always changing. “The types of scientists we learn about as we expand is growing. We’re always meeting with scientists to learn about the challenges they’re facing and how we can improve our user interface to accommodate their needs.” As the Head of User Experience, Matan works with his team to incorporate user feedback into new platform prototypes and create assets for the marketing team to use to promote their product to scientists. He says one of the most rewarding aspects of his role at BenchSci is “the satisfaction in designing solutions to complex scientific problems.”
The BenchSci office space
More than half of the employees at BenchSci are scientists and engineers. “The scientists understand what the platform needs, and the engineers know how to make it happen” Casandra said.
Of course, BenchSci could not exist without the last essential component of the team: Growth and Sales. As the Head of Academic Relations, Maurice Shen spends most of his time traveling to academic institutions, meeting scientists, and presenting demos of the platform. “The antibody crisis is a big deal, especially for young scientists who may not know how to select the right antibodies for their experiments,” Maurice explained. “We host seminars to talk about our platform, and how we are a solution to that problem.” As a direct result of this outreach, BenchSci is currently used at more than 2,000 academic institutions worldwide.
You may have seen BenchSci in the news last year when they received funding from Google’s artificial intelligence-focused venture fund, Gradient Ventures. Notably, they were the first Canadian company to do so. With this investment comes access to resources, increased credibility, and much-deserved attention.
So what’s in the future for BenchSci? Expansion. In the next few years, BenchSci is aiming to become a diverse reagent intelligence platform, assisting scientists in reagent choice, assay development and experimental design. BenchSci’s platform already helps researchers find reliable antibodies 24x faster than before, accelerating the pace of scientific research. Without a doubt, the scientific impact of BenchSci’s platform will continue to grow in years to come.
To learn more about BenchSci or make a free academic account, visit https://www.benchsci.com/.