Expanding genetic toolbox for heterologous protein expression in non-conventional yeasts using artificial intelligence.

Heterologous protein expression is the production of vital proteins for improving our food sources and creating life-saving drugs. However, predicting how much protein will be produced is difficult because we do not fully understand how genes are regulated and controlled. One way to produce proteins is by using certain types of yeast called Non-Conventional Yeasts (NCYs). These yeasts have traits that make them better at producing proteins, such as using complex carbon sources and being better suited genetically for producing specific products. However, we do not have enough information about these yeasts to use them effectively for protein production. To use NCYs for protein production, we need to develop tools to control gene expression in each host. We are using machine learning to help us predict how genes will be regulated and to make it easier to use NCYs for protein production. This will help us create sustainable, low-carbon, and cost-effective methods for producing proteins. We will test around 10,000 genes optimized for high protein expression in about 100 NCYs to see if we can control gene expression in these yeasts. This will help us understand gene expression and how it can be controlled, which could significantly impact biotechnology. Our ultimate goal is to use NCYs to produce proteins sustainably and cost-effectively.