BenchSci: An Open Access Platform to Search for Antibodies

Finding the right antibodies for a research is a process that can take up valuable time. BenchSci is a search engine with a machine learning algorithm that helps professionals with finding reliable antibodies and thus to speed up their studies. The platform translates closed and open access data into recommendations for specific experiments, decoding research papers to recognize used antibodies and the associated experimental contexts in each paper.

Scientists can search for their protein of interest on BenchSci. Results are then presented, showing individual figures from scientific papers in which antibodies against that protein were used. To narrow down their search, users can filter on the antibody use cases to find out what antibodies are best to use for their research.

“Machine learning is transforming biomedical research,” added Gradient Ventures founding partner Ankit Jain. “BenchSci’s technology (…) enables academic researchers to spend less time searching for antibodies and more time working on their experiments.”

BenchSci’s users have the option to either examine figures first and then identify the antibody used, or examine the antibodies first and then view their supported figures. It’s possible to identify figures and used antibodies by filters like technique, tissue and cell lines, and once the product or figure of interest has been found, a user is able to directly view the original paper or supplier website. Also, the website allows for users to save favorites, share figures or products with others and write reviews.

Two COMBACTE studies, COMBACTE-NET’s SAATELLITE and COMBACTE-MAGNET’s EVADE work with antibodies. Testing the safety, pharmacokinetic and pharmacodynamic characteristics, and the efficacy of monoclonal antibodies against S. aureus and P. aeruginosa.