Compound library design

Natural Product Scaffolds and Protein Structure Similarity Clustering (PSSC) as Inspiration Sources for Compound Library Design in Chemogenomics and Drug Development... 

PSSC as Guiding Principle for Compound Library Design 5.1. Protein structure similarity clustering (PSSC)... 

Figure 3. Approaches to cluster protein targets for screening of small molecule libraries. Currently, protein targets are clustered based on functional or amino acid sequence similarity (left). We propose to cluster proteins purely based on structural similarity of protein cores and to apply this clustering to compound library design (right).

Figure 4. Nature s conservatism and diversity found in a protein structure similarity cluster. The conservatism found in the cluster can be used for the selection of pre-validated guiding structures for compound library design. The chemical diversity of the compound library is needed to address the biological diversity found in the binding sites of the cluster members.

The PSSC concept provides a conceptually new principle to inspire compound library design for chemical biology and medicinal chemistry... 

Application of PSSC is not limited to existing crystal structures of proteins. It could be applied to structures derived from homology models as well. A homology model, as applied for the 11 HSD types 1 and 2 enzymes performed well in the first application of PSSC for compound library design. Furthermore, it should be noted that the concept has not been developed to make predictions on structural complementarity or the appropriate orientation of functional groups in the binding site. 

In Chapter 3, the group of Prof. Herbert Waldmann at the Max Plank Institute for Molecular Physiology provides a rationale on how natural products which play historically a predominant role in drug discovery are efficiently used in combination with protein structure similarity clustering to inspire directed compound library design and target identification. 

N. P. (2009) Compound library design for targeted families, in (Jacoby, E. ed.) Chemogenomics. Humana Press, New York, pp 21—46. 

Two examples Mobius from Coa-lesix, Inc. A GA driven, R-group based compound/library design tools (http //www.coalesix.com/FAQ.html) and C2-LibX from Accelrys, Inc. contains a GA based library design module (more from http //accelrys.com/ ). 

Compound libraries designed for the identification of new lead structures should not contain large numbers of closely related compounds (which are likely to have similar properties), but highly diverse compounds with widely different properties. Because compound libraries are prepared using the same reaction sequence for each... 

The identification of Gdk4-specific amino acid residues allowed us to focus on target residues that interact with substituents on lead compounds. Library design based on the locations of these residues and the binding modes of lead compounds enabled us to develop potent and selective compounds efficiently. Among them, compound 11 showed excellent selectivity not only over Cdkl/2 (780-fold/... 

In certain compound acquisition or screening projects, the selection of compounds may only be conducted at the plate level, where a whole plate of compounds is either selected or not selected. This is due to reasons of logistical considerations and concerns of compound stability during freeze-thaw cycles. Thus, compound library design or screening set design should take this constraint into consideration. ... 

---