De novo protein design has recently emerged as an attractive approach for studying the structure and function of proteins. This approach critically tests our understanding of the principles of protein folding; only in de novo design must one truly confront the issue of how to specify a protein's fold and function. If we truly understand proteins, it should be possible to design receptors, enzymes, and ion channels from scratch. Further, as this understanding evolves and is further refined, it should be possible to design proteins and biomimetic polymers with properties unprecedented in nature.

De novo protein design provides an attractive approach for the construction of models to probe the features required for function of complex metalloproteins. The metal-binding sites of many metalloproteins lie between multiple elements of secondary structure, inviting a retrostructural approach to constructing minimal models of their active sites. The backbone geometries comprising the metal-binding sites of zinc fingers, diiron proteins, and rubredoxins may be described to within approximately 1 Å rms deviation by using a simple geometric model with only six adjustable parameters. These geometric models provide excellent starting points for the design of metalloproteins, as illustrated in the construction of Due Ferro 1 (DF1), a minimal model for the GluXxx-Xxx-His class of dinuclear metalloproteins. This protein was synthesized and structurally characterized as the di-Zn(II) complex by x-ray crystallography, by using data that extend to 2.5 Å. This four-helix bundle protein is comprised of two noncovalently associated helix-loop-helix motifs. The dinuclear center is formed by two bridging Glu and two chelating Glu side chains, as well as two monodentate His ligands. The primary ligands are mostly buried in the protein interior, and their geometries are stabilized by a network of hydrogen bonds to second-shell ligands. In particular, a Tyr residue forms a hydrogen bond to a chelating Glu ligand, similar to a motif found in the diiron-containing R2 subunit of Escherichia coli ribonucleotide reductase and the ferritins. DF1 also binds cobalt and iron ions and should provide an attractive model for a variety of diiron proteins that use oxygen for processes including iron storage, radical formation, and hydrocarbon oxidation. P roteins use a limited repertoire of metal ion cofactors to help catalyze a multitude of reactions. For example, diiron sites (1-4) mediate reversible oxygen binding in hemerythrins, whereas they function as hydrolytic centers in phosphatases. Structurally similar diiron sites also mediate a number of oxygendependent oxidative processes. Ferritins serve as ferroxidases, while other diiron proteins catalyze hydroxylation, epoxidation, and desaturation reactions. Further, a diiron site in Escherichia coli ribonucleotide reductase is responsible for the formation of a Tyr radical. How do the structures of these proteins tune the chemical properties of a common diiron center to obtain such a diversity of highly specific catalysts? This question is being addressed through the study of the natural proteins as well as the study of small-molecule diiron complexes (1-4). Although impressive progress has been made on both fronts, these approaches have inherent limitations. The study of large proteins is hampered by their extreme complexity, and it is difficult to synthesize small-molecule models capable of simultaneously binding diiron, oxygen, and various substrates. Recently, we and others have sought a molecular middle ground between these two extremes through the design of small proteins and peptid...

Alchemical free energy calculations hold increasing promise as an aid to drug discovery efforts. However, applications of these techniques in discovery projects have been relatively few, partly because of the difficulty of planning and setting up calculations. Here, we introduce Lead Optimization Mapper, LOMAP, an automated algorithm to plan efficient relative free energy calculations between potential ligands within a substantial library of perhaps hundreds of compounds. In this approach, ligands are first grouped by structural similarity primarily based on the size of a (loosely defined) maximal common substructure, and then calculations are planned within and between sets of structurally related compounds. An emphasis is placed on ensuring that relative free energies can be obtained between any pair of compounds without combining the results of too many different relative free energy calculations (to avoid accumulation of error) and by providing some redundancy to allow for the possibility of error and consistency checking and provide some insight into when results can be expected to be unreliable. The algorithm is discussed in detail and a Python implementation, based on both Schrödinger's and OpenEye's APIs, has been made available freely under the BSD license.