“QUALITY IS NEVER AN ACCIDENT. IT IS ALWAYS THE RESULT OF INTELLIGENT EFFORT”. John Ruskin
Our Artificial Intelligence (AI) based systems, which encode deep rooted knowledge for compound design and assessment, can be applied comprehensively to conventional single-target drug discovery projects.
To identify those targets most likely to be chemically tractable, we first investigate the drugability of each opportunity, by which we mean the likelihood of a target to selectively bind a well-balanced small molecule.
A valuable component of this total process is to consider what range of chemistry and assay information are available for each target of interest as well as the extent to which similar chemical ideas associate with other unwanted targets.
Our approach was first published in Nature (2012); Automated design of ligands to polypharmacological profiles and has subsequently extended and refined.
Key aspects of the Design, Make, Test, Analyse cycle are described below.
PROGRESSING NEW AND NOVEL TARGETS
For high-interest targets, where patent, publication and proprietary data are insufficient to seed the evolutionary design algorithms, we apply a preliminary round of intelligent fragment screening, driven by Surface Plasmon Resonance (SPR) experiments.
The rapid application and high sensitivity of SPR decreases the delays often encountered with conventional assay development and permits fragment screening without the need for 3D structure. In our hands the approach can be productively applied to both soluble targets and challenging, high-value G Protein Coupled Receptors (GPCRs).