Profiling Entire Libraries

Figure 15.6a shows a chemical space representing the diversity of the entire library, on which the 13 active series have been highlighted. Figure 15.6b shows the same chemical space in which the points have been colored according to the scores of the corresponding compounds against the required property profile. 

In summary, examples have been provided in which both ligand-based and target-based methods performed decently when profiling compounds against the family of nuclear receptors. Therefore, it is reasonable to say that these methodologies have reached a sufficient level of maturity to be applied sensibly for designing the next generation of chemical libraries directed to entire protein families. 

Therapeutic antibodies, especially the latest generation platforms (humanized and fully human), have a potentially superior specificity due, for example, to the high number of variants in the selection process (more than 10 °) that cannot be matched by small-molecule libraries that show diversities in the range of 10 (see also Part V, Chapter 2). This often leads to better side-effect profiles shown in the better overall clinical survival rates, especially for humanized and fully human antibodies (e.g. [23]) (see also Part IV, Chapter 3). In addition, monoclonal antibodies have specific capabilities that are entirely unique (e.g., activation of immune response). 

---