Abstract:
Systems and methods for identifying enzymes for catalysing biochemical reactions include receiving input of reaction(s) and/or target molecule(s) along with data associated with chemical conversion, determining functional and linker region(s) in the input, scanning a transformation library for the determined functional region(s) of the reaction(s) and/or the target molecule(s) to find similar functional region(s) within the transformation library, assigning the reaction(s) and/or target molecule(s) to group(s) of the transformation library showing a high similarity to the transformation, computing a metabolite similarity score of the reaction(s) and/or target molecule(s) with respect to one or more reactions of the assigned group, and identifying enzyme(s) associated with the reaction(s) of the assigned group having a high metabolite similarity score. A transformation library is also generated.