Network-Driven Drug Discovery
Dr Ben Allen1, Dr Jonny Wray1, Dr Colin Stubberfield1, Dr Marie Weston1, Dr Victoria Flores1, Dr Adam Sardar1, Dr Sree Vadlamudi1, Dr Alan Whitmore1
1e-Therapeutics
Network pharmacology models cells as networks of interacting proteins. Within this paradigm, a disease state is identified as a disorder of the entire network, and the target for pharmacological intervention becomes a set of proteins. Network modelling can be used to identify a set of key proteins and their interactions which can be targeted to correct the network disorder in a disease setting.
e-Therapeutics proprietary Network-driven Drug Discovery (NDD) platform comprises a suite of powerful, custom computational tools that tap into large-scale proprietary databases, and employs data mining, machine learning, artificial intelligence, and network science to tackle complex diseases in an efficient and effective way. We have successfully implemented and validated a highly productive (NDD) approach to identify NCEs in diverse areas of biology. Our process is:
> Driven by Biology
We focus on the biology of disease processes aiming to identify small molecule hits.
We use disease specific expertise as well as computation methods to construct network disease models.
> Mechanistic – not black box/statistical modelling
We explicitly model the complex cellular mechanisms of the disease processes we aim to disrupt.
Mechanism is critical in guiding post-computational compound development in the lab.
> Multiple computational tools and data sets – not just AI
Our core approach utilizes biological network analysis and computational optimisation.
AI techniques are utilized to augment databases used for mechanistic modelling.
Both compound bioactivity and regulatory interactions are augmented to improve predictions and allow exploration of areas of chemistry and biology with less available experimental data.
> Validated
Output from the platform are drug-like, small molecules that are active in physiologically relevant tests.
Identified hits have been chemically optimized into leads in multiple projects from multiple therapeutics areas, including: oncology, immune oncology, autoimmune disorders, antivirals, CNS and immunology.
The presentation will summarise the key tools and methods that comprise the platform, with particular focus on the use of AI for data augmentation. We then demonstrate the validation of this approach by showing network active molecules that have been experimentally identified as hits in phenotypic assays, and taken through into hit and lead generation across multiple projects.
