One of the startup ideas I learned about at the Computational Biology Pitchathon was a project focused on using machine learning to help wet-lab teams prioritize which experiments to run next. The team’s concept was a software platform that takes in existing genomics or proteomics data plus a lab’s previous results and then suggests the most promising follow-up experiments, with an estimate of cost, time, and likelihood of success. Their goal was to reduce the amount of trial-and-error in early-stage biotech research so small labs and academic spinouts can use their limited funding more efficiently. I liked that the idea didn’t try to replace scientists, but instead acted like a “decision-support layer” that helps researchers decide where to focus their efforts.
Related Challenge:
challenge_points:
Image:
