Scinapsis Analytics Inc.

Published purpose

BenchSci enables pre-clinical pharmaceutical research scientists to select experimental reagents and model systems based on the documented history of experiments. As history contains a very large number of documented experiments, a “search” can result in a huge number of results. Many factors play into every drug development experiment, making the selection between all options difficult. Today, BenchSci uses a simple historic popularity ranking and this project will replace this ranking with a new ranking that incorporates as many experimental design factors as possible enabled by an ML model. This should create a significantly more accurate ranking without requiring an explicit understanding of every factor in a successful selection.