Data science is a branch of science for deriving knowledge from data. Practitioners of data science seek to uncover patterns in data to inform our understanding of measurable reality. A variety of tools ranging from classical statistical techniques to modern deep learning methods are applied to elucidate structure in data, build predictive models, and extract insights. Data science can be used to learn more about anything for which data exists, and here at Shiru, data science is foundational in our efforts to discover plant proteins for functional food ingredients.
Proteins have been a focus of scientific study since the 18th century (The Vegetable Proteins, 1909). Over the centuries since, vast amounts of individual data points have been generated about countless different aspects of proteins. We at Shiru use data science to study these individual data points in aggregate, leveraging the treasure trove of historically and contemporarily produced protein data to drive our search for functional food proteins. Working in conjunction with our laboratory scientists, who are constantly outputting new data about proteins via their efforts in strain engineering, precision fermentation, and high-throughput screening, we are able to iteratively supply our models with an ever expanding pool of ever more relevant data points. The impact of this is that we are able to continuously grow the capabilities of our discovery platform. In fact, at Shiru we are creating the first database that links protein identity with food functionality - made possible only by data science.