The invention provides efficient methods for combining single-substitution libraries of nucleic acids that span and encode proteins of interest and for selecting resultant mutant proteins after expression which have improved properties or characteristics. Specifically, the methods comprising synthesizing a single substitution library for each of a plurality of domains of a protein; expressing separately each member of each single substitution library as a pre-candidate protein; selecting members of each single substitution library which encode pre-candidate proteins which exhibit an improvement in the one or more predetermined characteristics to form a selected library; shuffling members or the selected libraries in a PCR to produce a combinatorial shuffled library; expressing members of the shuffled library as candidate proteins; and selecting mutant proteins which have improved properties or characteristics.