We examine the role of human- and data-driven approaches in deciding what to add and what to remove from a screening collection. Our use of parallelised fragment screening to evaluate target ligandability generated significant amounts of data, revealing characteristics relating to the promiscuity/privilege of fragments. Lessons learned from this and subsequent experiences in fragment hit followup will be described, along with steps that have been taken to improve future prospects. The range of commercial analogues available around a set of hits is a key factor determining the selection, speed and success in the progression of hits into synthetic optimisation. In particular we describe our approach to identifying fragments that have a reasonable range of analogues with single-point modifications in multiple positions. This enables informative judgment of the SAR around the hits beyond the constraints of structural information. An outlook on other key factors surrounding project progression will also be presented, including the challenge equating ligandability and druggability in antibacterial targets.
To view the video assoicated with this lecture