Cherry picking, combinatorial libraries

Fig. 15.24 RADDAR a nearly infinite number of chemical structures can be generated by computationally enumerating combinatorial library proposals. Cherry picking of interesting candidates (proposed structures) based on defined computational algorithms allows the chemist to synthesize the relevant subsets of theoretically accessible molecules.

Product-based approaches can be divided into those that take the combinatorial constraint into account such that each reactant in one pool appears in a product with every reactant from every other reactant pool, and those that merely pick product molecules without consideration of the synthetic constraint. The latter approach is often referred to as cherry-picking and is synthetically inefficient as far as combinatorial synthesis is concerned. In this chapter the emphasis is on product-based library design methods that take the combinatorial constraint into account. 

In product-based selection, the properties of the resulting product molecules are taken into account when selecting the reactants. Typically this is done by enumerating the entire virtual library that could potentially be made. Any of the subset selection methods described previously could be used to select a diverse subset of products, however the resulting subset is very unlikely to represent a combinatorial subset. This process is known as cherry-picking and is synthetically inefficient as far as combinatorial synthesis is concerned. Synthetic efficiency is maximized by taking the combinatorial... 

Any of the compound selection methods that have been developed for reactant selection can also be applied to the product library in a process known as cherry picking. A subset library selected in this way is shown by the shaded elements of the matrix in figure 3. However, a subset of products selected in this way is very unlikely to be a combinatorial library (the compounds in a combinatorial library are the result of combining all of the reactants available in one pool with all of the reactants in all the other pools). Hence, cherry picking is combinatorially inefficient as shown in figure 3 where 7 reactants are required to make the 4 products shown. 

Although most applications were of the cherry-picking type design, the combinatorial design of new chemical libraries should also be feasible. In this case, the scores obtained with the various models can be used to sort the virtual library, followed by building block frequency analysis cf. Focus2D) to determine which reagents should be used in chemical synthesis. Alternatively, combinatorial optimization approaches, such as those in described in ref. 4, can be applied where the model-predicted scores are used as the objective function for optimization. 

The chapter begins with a discussion of similarity and diversity measures and how they can be applied in a virtual screening context. The various computational filters in use are also discussed. The rest of the chapter is concerned with different approaches to combinatorial library design, beginning with reagent-based methods followed by product-based approaches of cherry picking and combinatorial subset selection. Finally, approaches to designing libraries optimized on multiple properties simultaneously are discussed. 

Figure 4 A two-component combinatorial library is represented by a 2D array, (a) A cherry-picked subset of nine compounds is highlighted, (b) A 3x3 combinatorial subset is highlighted.

MCRs can be considered the cradle of combinatorial chemistry [56]. Other than the stepwise library synthesis described in the library design section above, MCRs have the advantage of the very short reaction sequence. This allows for the possibility to apply extreme cherry-picking of the desired products, resulting in very low matrix coverage and high structural diversity regarding the decoration of the scaffold. 

---