Systems and methods for building a comprehensive genome scale metabolic model. The method includes determining that at least one of a hypothetical profile annotation and an uncharacterized profile annotation is available in a profile annotation associated with a protein. Further, the method includes performing a machine learning procedure on at least one of the hypothetical profile annotation and the uncharacterized hypothetical profile annotation after a fuzzy string matching and ranking procedure. Further, the method includes obtaining all the possible protein annotations based on a fuzzy string matching procedure. Further, the method includes performing a rank procedure on at least one of the hypothetical profile annotation and the uncharacterized profile alternative annotation. Further, the method includes identifying all possible metabolic reactions for the protein annotation obtained by the fuzzy string matching procedure. Further, the method includes ranking the protein annotation based on the metabolic reactions.