Progenus, an innovative company specializing in genetic testing, focuses on DNA research and development (R&D) in fields like animal husbandry, agriculture, and the food industry. As a leader in the sector, Progenus is consistently seeking ways to utilize cutting edge technologies to enhance their operations and R&D efforts. We assisted Progenus in tackling a key challenge related to protein-DNA binding. The traditional method of conducting systematic laboratory tests for this purpose was found to be both labor-intensive and resource-heavy. Progenus was in need of a more streamlined solution that would facilitate the process of identifying the DNA sequences most likely to bind to a specific protein. Implementing such a solution would significantly decrease the resources needed for the design and development of new PCR tests.

Despite the limited amount of available data, our team of experts developed a custom artificial intelligence solution capable of filtering a subset of ''high quality'' sequences for experimental testing. Our results lead to an eight-fold decrease in the number of tests required to determine optimal DNA/RNA sequences and the results were validated on multiple proteins. We achieved this by leveraging custom machine learning techniques and an innovative approach to engineer useful features from raw DNA data, significantly enhancing the probability of suggesting experimentally viable recommendations compared to standard selection techniques.

Our innovative AI-driven solution substantially improves the accuracy of Progenus' DNA-protein binding pair recommendation system. The solution saves time and resources, offering significant operational improvement to Progenus, and a pathway towards more efficient and cheaper genetic tests.