Commentary  |
AI & Machine Learning

AI and ML for Target Validation & Lead Optimization

Edited by Oliver Picken |
07 June 2022
Jackie Hunter, Board Director at BenevolentAI, shares details on her company's unique approach to target Identification and lead optimisation.

This article presents an updated overview of Jackie Hunter’s (Board Director, BenevolentAI) presentation during our Pharma IT 2020 event. Article edited by Oliver Picken.  

Artificial intelligence (AI) and machine learning (ML) are having a significant impact on the entirety of drug discovery and development. The current process is extremely challenging and time-consuming. It costs an average investment of USD 2.6 billion to bring a drug through research and development to the market. The whole process often takes at least ten to fifteen years, with 96% of new drug programmes ending in failure.

Advanced new technologies have the power to revolutionise the drug discovery process, particularly during target discovery & validation. AI and ML have the potential to radically change how we look at what to target, what hypotheses we generate, and how we understand disease processes.

BenevolentAI operates at the forefront of this emerging AI drug discovery sector. We have a track record of success, pioneering a fundamentally differentiated and validated approach, as evidenced by the fact that our AI discovery platform has generated over 20 disease programmes in our pipeline. Formed with a world-class leadership team and experts in the fields of drug discovery, AI, engineering, informatics & product development, our goal is to tackle the intractable challenges of drug discovery, including the high cost, failure rates, and time to market. We aim to better understand disease biology and develop more effective therapies for millions of patients worldwide.

Target Identification & Validation

The BenevolentAI Platform™ is built upon four key development pillars, the first being knowledge foundations. Our Knowledge Graph integrates the vast amount of biomedical data, research, and literature at scale and serves as a data engine for our end-to-end Platform and drug discovery programmes. It is impossible for scientists to effectively process, mine and exploit the potential of this ever-growing body of biomedical data and research. All of our AI and ML models and workflows are supported by our Knowledge Graph, which serves as a data engine for our end-to-end platform and drug discovery programmes. It is created from a mix of publicly available biomedical data mined from structured sources, information extracted from scientific literature and internally generated experimental findings. It is continuously updated with proprietary insights and new knowledge generated from our experiments and AI models and supplies a complete and unbiased representation of biomedical data.

This approach does not rely on any one data type or experimental methodology – we purposefully aim to be more ambitious and comprehensive.  We have developed AI methods that operate on top of our Knowledge Graph to extract and infer unique biological insights entirely proprietary to BenevolentAI. Our exploratory tools and predictive models empower scientists to explore relationships in the graph between biological entities and disease networks.

Secondly, we have developed a unique approach to Target Identification. Target identification and characterisation begins with identifying the function of a possible therapeutic target and its role in the disease. A good target should be efficacious, safe, meet clinical and commercial requirements and be druggable. They are often a particular protein, but can also be a chemical signalling pathway, small molecule or DNA binding site, among others. Target identification is the first step in developing a cure for a given disease. Unfortunately, one of the primary reasons drugs fail in clinical trials is that the wrong target is selected from the outset.

Our ML infrastructure powers large scale predictions for disease targets. Our tools enable scientists to run in silico experiments in real-time, yielding results that guide the discovery process and ultimately identify novel biological targets. To do this we integrate a wide range of data modalities that represent biomedical knowledge captured in scientific databases and literature Importantly these data include genomic and genetic information derived from patient tissue. Harmonising all these data together allows them to be analysed much more effectively using our AI and ML tools.

We factor in Precision Medicine at the earliest stages of drug discovery, leveraging multimodal patient and endotype-specific data, i.e., distinct disease entities each defined by a specific biological mechanism. Our precision medicine workflows empower drug discoverers to identify patient subgroups that will likely respond to a particular treatment and inform the design of clinical trials.

We have a fully equipped wet lab in Cambridge in the UK with biology, chemistry, CMC and DMPK, and we have in-house CRISPR, RNA seq and human iPSC capabilities. Experimental data from the lab and disease-relevant expression data are integrated back into our Knowledge Graph to further enrich our representation of human biology.

Final Thoughts & Conclusion

BenevolentAI’s platform is scientifically and commercially validated and has already produced a rich pipeline of over 20 platform-generated disease programmes.

Our partnership with AstraZeneca has grown from strength to strength.  The initial partnership was focused on identifying novel targets for Chronic Kidney Disease and Idiopathic Pulmonary Fibrosis. We have already had three novel targets selected by AstraZeneca for entry into their portfolio. We have recently expanded this collaboration into two new disease areas – systemic lupus and heart failure.

Another significant partnership we have developed is with Eli Lilly. Our work began when we uncovered that their drug, baricitinib, approved for rheumatoid arthritis, also has previously unknown antiviral effects, which means it can inhibit viral entry into cells. The drug has been shown to reduce mortality in hospitalised patients with COVID-19 by 38%, and it has been approved by the FDA to treat COVID-19.

AI and ML are proving to be valuable tools in drug discovery, and the potential to reduce the cost of new drugs is of a high interest to the pharmaceutical industry. As a result, a great deal of effort and funding is being spent on new tools and methods. BenevolentAI is one of the leading companies in this area and are at the forefront of exploring how AI and ML can be applied across the entire drug discovery process, including target validation and lead optimisation.

For more on innovative uses of AI, consider joining us for Pharma Data UK: In-Person, taking place in London on 08 – 09 September 2022.

Speaker Biographies

Jackie Hunter – Board Director, BenevolentAI

Jackie Hunter is a Board Director of BenevolentAI. Jackie has over thirty years of experience in the bioscience research sector, working across academia and industry including leading neurology and gastrointestinal drug discovery and early clinical development for GlaxoSmithKline. She founded OI Pharma Partners in 2010 to support the life science sector in harnessing the power of open innovation and most recently was Chief Executive of the Biotechnology and Biological Sciences Research Council. She holds personal chairs from St George’s Hospital Medical School and Imperial College and serves on numerous Advisory Bodies and Boards including the Technology Advisory Board of BP plc and A*Star Board Singapore. She also Chairs the Boards of Brainomix Ltd, Stevenage Bioscience Catalyst and the Sainsbury Laboratories Norwich.

Share this article

Share on facebook
Share on twitter
Share on linkedin

Sign up for our monthly PharmaTec Newsletter

You may also be interested in...

Insight Article
Digital tools to boost productivity and optimise pharma manufacturing are becoming increasingly prevalent through smart sensors, AI, ML, virtual reality, and cloud computing. Digital transformation is here for Pharma, and it is changing companies' priorities towards digital solutions.  
04 February 2022
Workshop Speakers Include: Brian Martin, Head of AI in R&D Information Research, Senior Principal Data Scientist, AbbVie | Kevin Hua, Senior Manager AI/ML, Bayer U.S. LLC | Peter Henstock, Machine Learning and AI Technical Lead, Pfizer | Thibaud Coroller, Principal Statistical Consultant, Novartis
08 July 2020

Continue browsing

Share this article

Share on facebook
Share on twitter
Share on linkedin

Join our PharmaTec mailing list

We produce cutting edge congresses and summits for the Life Sciences Industry, bringing together industry leaders and solution providers at a senior level, creating the opportunity to partner, network and knowledge share.

Contact Us:

Copyright Oxford Global Marketing Limited. All rights reserved.

Member Community Login

Stay up to date

Sign up for our monthly Editorial Newsletter to keep up with all things PharmaTec