All life processes, either normal or pathologic, manifest through more or less specific protein-protein interactions (PPIs). The more detailed information we have on these PPIs, the better we understand their mechanisms, which also opens the possibility of mechanism based intervention. Deciphering the structural bases of any given PPI requires large numbers of protein variants. Moreover, understanding the physiologic or pathologic roles of a given PPI requires ultra-selective compounds that can bind the corresponding interfaces and block the original PPI. Directed protein evolution is an ideal framework for both approaches. In the frame of directed evolution, in a precisely designed manner we produce billions of variants of the protein we aim to study or further develop. Then, from this library we select variants that either a) maintain their original function, - and this way we reveal the structural elements essential for the original function -, or b) select for some entirely new binding specificity for selective blocking a PPI. This latter approach could lead to developing biological pharmaceutics if blocking the original PPI provides some therapeutic solution.