Ligand-based de novo Design Techniques

De novo drug design methods have been developed almost entirely from a structure-based design perspective where structural co-ordinate information about the site is available. However, in the absence of target structural data, design can only proceed by using known actives found from screening campaigns. This is a very common problem in the pharmaceutical industry; most novel targets lack structural data. If hits from screening have been found, can they be modified to provide a good IP position? An extension of this problem arises in the focused design of compound libraries for a family of genomically related targets.

If de novo design techniques are to be used in ligand-based design, models of the site have to be constructed from the available ligand data and validated to provide the best models. If this can be achieved, the models can be used in a de novo structure generation procedure. This presentation will illustrate how ligand-based de novo design can be achieved and applied to a family of aminergic GPCR targets.