A Promising Advance in AI-Enabled Drug Discovery

First lead candidate identified with the help of the AI-powered Logica platform

Bringing new medicines to patients is an exceptionally difficult and lengthy journey fraught with many pitfalls, and only a handful of candidates make it to market. This high level of attrition means that drug discovery scientists can consider themselves exceptionally fortunate to work on projects in the discovery phase that result in a medicine reaching patients. Biology is hugely complex and the more we advance our knowledge of science the more we confront a counterbalance by tackling ever more challenging diseases and molecular targets.

There is always talk of the latest technology and how it will revolutionise what we do, with examples including Computer-Aided Drug Design (CADD) in the 1980’s and combinatorial chemistry in the 1990’s. When the excitement subsides, these technologies find their place in the drug discovery toolbox and the true impact becomes apparent. But the harsh reality is that no magic solution can make drug discovery easy.

Artificial Intelligence in Drug Discovery

AI is becoming a daily reality, and that reach has extended to science. In drug discovery, there is a recent trend where AI is being used as an umbrella term for all things predictive and generative. AI is expected to have a significant drug discovery impact and at times has been portrayed as the next great technology to transform drug discovery. All of which seems to drive an obsession on whether AI is discovering drugs. 

The truth is more complex. AI will certainly play a role in drug discovery, alongside existing computational and experimental science, and most importantly the drug discovery experts. AI is evolving how we think about drug discovery, challenging scientists to fully utilise the wealth of data from past and present projects to guide how targets are selected, how initial hit compounds are found that modulate the targets, and how these are optimised through to preclinical candidate compounds ready for safety studies. In this way, AI is transformational in drug discovery and AI enabled drug discovery platforms will simply become embedded in our drug discovery culture.

Lead Candidates from Logica

A recent example of success has been the announcement of the first Logica lead candidate compound in a previously disclosed partnership with Flagship’s Pioneering Medicines. Logica - a partnership between Charles River and Valo Health - is delivering and democratising AI enabled small molecule drug discovery to partner organisations. The significance of this first success is the demonstration of an early proof-of-concept for the platform and its implementation. As with any drug discovery program there are scientific challenges on the journey, and what we learn from this first Logica success will enable fine tuning of the platform. 

How Does AI in Drug Discovery Optimisation Work?

The success of AI enablement in small molecule optimisation phases is tethered to the quality of models and how they are deployed in Design Make Test Analyse (DMTA) cycles. Intentional homogeneous data generation will enhance the domain of applicability and performance of predictive models for ADME and early safety end points across the portfolio of projects. Individual projects will also benefit through deployment of localised models for prediction of project specific end points including on-target potency. 

These localised models will develop throughout the optimisation cycles in line with data accumulation for the chemical series of interest. Adoption of a prediction driven culture and the interleaving with physics-based computational chemistry is imperative to maximise value from all computational approaches to drug discovery optimisation phases. Platforms such as Logica will continue to reap the benefits in these areas as the technology and deployment matures through an increasing portfolio of projects.

Small molecular drug discovery is continuously evolving, and AI enablement can work together with physics-based computation and enhanced automation to increase efficiencies and shorten discovery phase timelines. A challenging but bright future awaits for the discovery and development of small molecule therapeutics.

eBook: "Translational Strategies Accelerate Drug Discovery"
This eBook focuses on translational science and discovery partnerships that advance novel therapies through a collective momentum, leveraging efficiencies in drug discovery to accelerate nonclinical research and increase clinical success.
Download the eBook