Platform Technology

Sosei Heptares boosts Artificial Intelligence and Computer-Aided Drug Design capabilities focused on GPCR-targeted drug discovery with new partnered initiatives

By Sosei Heptares | Oct 20, 2021

Sosei Heptares has recently presented progress in the development of novel approaches in combining AI molecule generation and GPCR Structure-Based Drug Design (SBDD) and will be participating in new collaborative initiatives in training in AI and Computer-Aided Drug Discovery (CADD).

Progress on the use of chemical language models for de novo design of GPCR ligands was recently presented by Morgan Thomas at the Royal Society of Chemistry’s AI in Chemistry 1 and European Research Network on Signal Transduction (ERNEST) 2 GPCR meetings, as part of a collaborative project between Sosei Heptares and the University of Cambridge.

During these presentations, Morgan described new ways to integrate AI methods in customized GPCR SBDD/CADD protocols 3 developed within the context of his PhD project, which is being supervised by Dr. Chris de Graaf (Head of Computational Chemistry at Sosei Heptares) and Dr. Andreas Bender (University of Cambridge).

Dr. Chris de Graaf indicated ‘This pioneering work highlights the value of SBDD scoring approaches to guide the exploration of novel chemical space with reinforcement learning and other AI algorithms. The collaborative project and experiences further strengthen the in-house and partnered AI augmented drug discovery efforts within Sosei Heptares, including amongst others, the AI-driven GPCR drug discovery technology collaboration with PharmEnable 4.’

Sosei Heptares will also be participating in novel initiatives in AI and CADD research training and method development, including the European Innovative Training Network (ITN) on Allostery in Drug Discovery (ALLODD), and a BBSRC Collaborative Training Partnership (CTP) on Artificial Intelligence in Drug Discovery.

In ALLODD 5, Sosei Heptares and 12 other European academic and industrial partners will train researchers in the development and application of novel interdisciplinary computational and experimental approaches in the discovery of small molecule allosteric modulators of proteins and understanding of their allosteric mechanism.


‘The exciting era of GPCR Structural Biology, including a plethora of GPCR structures determined using Sosei Heptares StaR technology, is offering unprecedented insights into allosteric ligand binding and modulation of receptor function. The ALLODD network provides a great opportunity to combine front running collaborative science with joint cross academic-industrial training of a new generation of drug discovery scientists,’ Chris de Graaf emphasised.


In the BBSRC CTP 6, Sosei Heptares will train PhD students in the development and application of novel AI drug discovery technologies, in the areas of cheminformatics, chemogenomics, and bioinformatics, together with Queens Mary University London, Exscientia and MSD.

“These and other initiatives help us to further strengthen Sosei Heptares’ AI augmented drug discovery capabilities. The combination of public and Sosei Heptares proprietary GPCR information enable us to use AI approaches to identify novel patterns and relationships between GPCR biological, chemical and structural data that we can use to supercharge GPCR structure-based drug discovery” added Dr. Chris de Graaf.



  1. AI in Chemistry Symposium View Event  
  2. 05 October - ERNEST Bari 2021 (
  3. Thomas, M., Smith, R.T., O’Boyle, N.M. et al. Comparison of structure- and ligand-based scoring functions for deep generative models: a GPCR case study. J Cheminform 13, 39 (2021). View Publication
  4. PhrmaEnable  View News Release
  5. ALLODD 
  6. Bio-researchers of the future get £22.5 million boost – UKRI  View News Release