Leveraging machine learning for the systematic optimization of protein function

Current gene synthesis technologies allow precise control of all sequence features critical for biological function. Variables at the levels of amino acid substitutions, vector elements and host genome mutations can all be critical for achieving the desired output. Modern machine learning tools and algorithms are well equipped to identify and quantify the relative contributions of each variable. We will describe the combination of gene synthesis and megadimensional informatics for the systematic engineering of biological function. Several unpublished case studies will be presented.