Three decades of structure- and property-based molecular design

Roche has pioneered structure- and property-based molecular design to drug discovery. While this is an ongoing development, the past three decades feature key events that have revolutionized the way drug discovery is conducted in Big Pharma industry. It has been a great privilege to have been involved in this transformation process, to have been able to collaborate with, direct, guide, or simply encourage outstanding experts in various disciplines to build and further develop what has become a major pillar of modern small-molecule drug discovery. This article is an account of major events that took place since the early decision of Roche to implement computer-assisted molecular modeling 32 years ago and is devoted to the key players involved. It highlights the internal build-up of structural biology, with protein X-ray structure determination at its core, and the early setup of bioinformatics. It describes the strategic shift to large compound libraries and high-throughput screening with the development of novel compound storage and ultra-high-throughput screening facilities, as well as the strategic return to focused screening of small motif-based compound libraries. These developments were accompanied by the rise of miniaturized parallel compound property analytics which resulted in a major paradigm shift in medicinal chemistry from linear to multi-dimensional lead optimization. The rapid growth of huge collections of property data stimulated the development of various novel data mining concepts with 'matched molecular pair' analysis and novel variants thereof playing crucial roles. As compound properties got more prominent in molecular design, exploration of specific structural motifs for property modulation became a research activity complementary to target-oriented medicinal chemistry. The exploration of oxetane is given as an example. For the sake of brevity, this account cannot detail all further developments that have taken place in each individual area of structure- and property-based drug discovery and it can only hint at important developments in other disciplines that have equally contributed to major paradigm shifts in Roche's small-molecule drug discovery efforts.