The present invention provides methods for efficiently generating and screening
protein libraries for optimal proteins with desired biological functions, such as improved binding affinity for biologically and / or therapeutically important target molecules. The method is performed
in silico in a high-
throughput fashion by mining the ever-expanding
database of
protein sequences in all living things, especially humans. In one embodiment, a method of constructing a designer
protein library comprises the steps of: providing an
amino acid sequence derived from a lead protein, referred to as a lead sequence; comparing the lead sequence with a plurality of
test protein sequences; Select at least two
peptide fragments having at least 15%
sequence identity with the leader sequence from each
test protein sequence, and the selected
peptide fragments form a selection
library; a
library of designed proteins is formed by replacing the leader sequence with the selection library. Libraries of designed proteins can be expressed
in vitro or
in vivo to generate libraries of recombinant proteins that can be screened for new or improved functions relative to lead proteins, such as antibodies against a therapeutically important target.