Revolutionising Drug Discovery and Development with Artificial Intelligence

As the global pharma industry continues its shift towards precision medicine in the quest for targeting undruggable targets, research and development (R&D) activities have experienced a tremendous surge, there by requiring cutting edge technologies that not only streamline drug development processes but most importantly manage costs. As a result, adoption of newer technologies such as Artificial Intelligence (AI)/ machine learning (ML), Large Language Models (LLMs) and Generative AI (a subset of the AI metaverse) are on the rise providing time and cost-efficient alternatives to manual processes that took years to complete. 

As partnerships between pharmaceutical companies and technology providers grow, newer technologies like Generative AI for de novo drug design and disease modelling, synthetic control arms and digital twins for clinical trials’ control groups and patient data, and many more are gaining traction, supporting all aspects of drug discovery and development.

Before venturing in detail about the many benefits of technology in drug development, it is important to understand the R&D landscape. In terms of the R&D expenditure, industry giants including Roche, Pfizer, JNJ, Merck, and Novartis take the lead with the highest expenditure in value. These companies are also among the top 5 with number of R&D projects, with a significant focus on oncology. Apart from oncology, neurology (CNS disorders), infectious diseases, and immunology are the next key focus areas across the industry. Infectious diseases, especially, gained attention during the pandemic, with several trials initiated to support the development of COVID-19 therapies. Antibiotics are also gaining traction, with a focus on developing therapies for antimicrobial-resistant bacteria.

Interestingly, large pharma companies are now reporting comparatively smaller pipelines than previous years, except for a few larger conglomerates such as Novartis, BMS, Pfizer, Eli Lilly, and Boehringer Ingelheim. This suggests that the contribution of top pharma companies to the growing pipeline is declining. Instead, smaller, and emerging biopharma companies with less than 10 products in the pipeline account for more than 60% of the pipeline.  This shift is what is driving a greater reliance on technology in the form of AI/ ML and generative AI. 

AI enabled Drug Discovery – A step ahead in Pharma innovation

One of the most promising applications of AI in drug discovery is in the early stages of identifying potential drug candidates which is simply called ‘target identification’. Traditional drug discovery involves screening millions of compounds manually in lab to find ones that are effective against a particular target, such as a protein associated with a disease progression. This process is not only time-consuming and costly, also results in large number of potential compounds failing to make it through the screening process.

AI offers a more efficient approach by using machine learning algorithms to analyze vast amounts of data and predict the likelihood that a given compound will be effective against a specific target. These algorithms consider a wide range of factors, including the chemical structure of the compound, its biological composition, its similarity to known drugs, and its predicted interactions with the protein of interest.

Alongside target identification, another area where AI is making significant strides is in predicting the efficacy and safety of drugs. Traditional methods for predicting drug efficacy rely on animal studies and clinical trials, which can be time-consuming and may not always accurately predict how a drug will perform in humans. Moreover, with FDA’s Drug Modernization Act 2.0 in action, use of computational models is emerging as a necessity to avoid animal testing. AI offers a more accurate and efficient alternative by creating 3D animal models, analysing large datasets of biological and chemical data and even analyse gene expressions to predict how a drug will behave in the human body. For example, DeepMind’s Enformer architecture improves the ability to predict how DNA sequence influences gene expression, there by advancing genetic research. 

Therefore, AI application in drug discovery and preclinical research has become prominent in the past decade, with companies like Recursion, Insilico Medicine, AbCellera Biologics, Schrodinger and Exscientia leading the way. Moreover, as the market prospers, these companies have evolved from being technology vendors to establishing themselves as AI-powered therapeutics providers. They are creating their own pipelines with suitable dry and wet lab capabilities, there by competing directly with pharma/ biopharma companies. 

Whilst majority AI vendors have re-configured their business models, some of these players continue to operate as Pure play AI vendors, where in in-house developed pre-clinical molecules are outlicensed to pharma companies for co-development agreements.  As a result, focus of AI in drug discovery is shifting towards prescriptive and predictive insights, aligning with the industry's shift towards precision health. Expansion is expected across providing outcomes-based solutions with key objectives including drug repurposing, target identification, drug screening, preclinical testing, and more.

AI-Enabled Clinical Trials – Streamlining Patient Access, Trials Protocols and Patient Recruitment

Post the pandemic, the industry is becoming more open-minded about the adoption of technology to support remote/virtual trial modalities. Clinical trials are a critical step in the drug development process, but they are also time-consuming and expensive. In-line with the aim of achieving patient-centricity, AI-enabled digital infrastructure can significantly reduce human errors in clinical trials by incorporating a self-learning system to improve outcome predictions and proactively deliver valuable insights to sponsors and CROs, imbibing improved trial designs and workflows, whilst also greater willingness amongst patients to participate in trials and ensuring adherence and retention. 
AI can support three critical aspects of clinical development:

1. Enable continuous capture of Real-World Data, which undergoes seamless cleaning, structuring, aggregation, coding, and storing in cloud-enabled platforms, providing easy access to stakeholders involved in the trial. 
2. Target some of the most crucial trial stages, associated recruitment, and monitoring, which are the critical pain points directly impacting trial success/failure.
3. Utilise cutting-edge modalities, such as ML and NLP, to capture, structure, and analyse patient-related data during clinical trials.

For example, Owkin’s federated learning-based AI application, leverage its robust data network from leading hospitals and healthcare centers. Also, the Owkin Connect framework allows data owners from disparate sources to authorise data access and track usage in real time, there by allowing sponsors to design the trial with the best possible clinical outcomes, identify optimal dosages as well as predicting the potential patient outcomes, based on previous clinical trial data. 

Therefore, as opposed to the traditional clinical trial processes, AI-enabled platforms have consistently improved therapeutic outcomes owing to improved recruitment and better patient engagement.

What is driving growth? 

Venture Capital Funding, Mergers and Acquisitions (M&A) and Partnerships have been the key driving factors of the AI industry.  Certara acquiring Vyasa Analytics, and Recursion acquiring Valence and Cyclica exemplify the many strides taken by industry leaders in further expanding their portfolios and retaining market leadership. On the other hand, Technology partnerships between clinical trial vendors and AI vendors are also not unheard of with players such as CRL and Valo Health's launching an AI-enabled drug discovery platform Logica™ supporting drug discovery, and Syneos Health's strategic agreement with Microsoft to accelerate AI across its clinical and commercial operations, have been driving market growth while expanding the vendor ecosystem.

Most importantly, Venture Capital (VC) funding has been significant, with players such as Recursion, Relay Therapeutics, Exscientia, BenevolentAI and many more receiving over $300-$400 million funding till date within the drug discovery spectrum and players in the clinical trials segment like ConcertAI reaching unicorn status. These investments are driving innovation in AI-enabled drug discovery and development landscape.

Outlook->

As the drug discovery and clinical development landscape continues to evolve, Artificial Intelligence is poised to play a critical role in realising the full potential of precision medicine, fostering improved outcomes, enhanced patient experiences, and greater efficiency in drug research. AI has been phenomenal in driving breakthroughs across lead optimisation, ADME/Tox modeling, genotypic and phenotypic screening analysis, amongst other areas there by transforming the way one ever imagined pharma innovation. 

With partnerships, M&As and VC funding supporting growth, these technologies are set to break new barriers while unlocking potential unreachable targets and there by accelerating drug development in a cost-efficient manner.