Chemistry & AI
The Emerging Comradeship
Ashok Kumar, President R&D, Ipca Labs
This article explores the transformative impact of artificial intelligence (AI) in chemistry and healthcare. It highlights how AI, through innovations like DeepMind's AlphaFold and advanced tools, enhances drug discovery, improves cancer detection, and optimizes research. Emphasizing collaboration, it suggests that chemists skilled in AI will lead future advancements while maintaining human creativity.
Disruptive technologies have consistently played a pivotal role in reshaping various aspects of our lives, leading to significant improvements in efficiency and effectiveness. Among these transformative innovations, artificial intelligence (AI) and machine learning (ML) have garnered considerable attention, fundamentally revolutionising computational systems.
A prime example of this transformative potential is the advent of ChatGPT, a sophisticated large language model (LLM) driven by generative AI. The latest iteration, ChatGPT-4 Omni, which debuted in May 2024, represents a remarkable leap forward. With billions of parameters, it seamlessly integrates text, audio, and emotional cues, enabling real-time communication with a human-like fluency that highlights AI’s evolving capabilities. OpenAI's upcoming model promises not only to conduct deep research and tackle complex science and math questions but also to exhibit human-like reasoning skills.
The integration of AI into chemistry, drug discovery, and healthcare is no longer a futuristic vision; it is a present-day reality reshaping how chemists conduct research. This shift is leading to more efficient drug development, personalised healthcare solutions, and a better understanding of human health, as illustrated by the following examples.
AI in Healthcare:
The impact of AI on healthcare has already been profound. For instance, CHIEF, an AI model, can detect up to 19 different types of cancer, predict patient outcomes, and identify the origins of cancer. Additionally, it recognises gene and DNA patterns associated with treatment responses, providing insights into the mutation statuses of oncogenes and tumor suppressors. This tool can greatly assist clinicians in evaluating and treating patients more effectively and in a timely manner.
By processing and analysing large, complex datasets to uncover patterns invisible to the human eye, AI has proven to be a game changer in the healthcare industry, essential for accurately diagnosing illnesses and determining optimal treatment pathways.
AI in Drug Discovery and Development:
Traditionally, it was believed that creative and innovative fields, including the arts and sciences, were exclusive to humans and would remain untouched by AI. However, recent developments in chemistry and related fields challenge this notion.
Discovering new therapeutic targets is critical yet requires a complex understanding of protein structures and behaviors. The DeepMind algorithm AlphaFold has significantly advanced our ability to predict the three-dimensional (3D) structures of proteins from their amino acid sequences. In July 2022, it unveiled the projected structures of over 200 million known proteins, drastically reducing the time required for this process from years to mere hours.
For instance, when the structure of a protein is unknown, AlphaFold can be combined with other AI tools, such as Pandamics and Chemistry 42, to facilitate new discoveries. The target discovery engine PandOmics clarifies the relationship between a target and a disease, while Chemistry 42 identifies potential binding sites for drug candidates by analysing AlphaFold's predicted protein structures. AlphaFold2’s ability to anticipate multiple protein configurations not only aids in understanding protein dynamics but also revolutionises drug development by identifying new targets.
A notable example is the creation of the Tyrosine Kinase-2 inhibitor (TAK-279), which has shown promising results in Phase 2b studies. The acquisition of TAK-279 by Takeda from Nimbus Therapeutics for $6 billion in 2023 underscores the tangible impact AI is having on pharmaceutical development. This success has instilled confidence in small biotech firms, prompting a shift from traditional medicinal chemists to AI-driven predictive software. As Akinsanya, President of R&D in Computational Chemistry at Schrödinger, stated, “We are seeing many biotech companies opting to acquire AI platforms rather than building large chemistry teams, allowing them to compete effectively with major pharmaceutical companies.”
According to CPHI’s annual report, based on data from 250 international pharma companies, more than half of all approved medications are expected to utilise AI in their development or production within the next decade. AI is also being extensively used in new drug development, optimising leads and predicting pharmacokinetics (ADMET) for potential candidates. Furthermore, AI now plays a crucial role in clinical trial patient subgroup selection and trial predictions, significantly reducing costs, time, and late-stage failures—essential in an industry where bringing a new medication to market can cost billions of dollars and take 12 to 15 years.
AI in Chemistry Research:
In chemistry research, AI’s potential is equally transformative. This field often involves exploring various methods for molecule optimisation and discovering compounds that meet multiple criteria. AI's ability to process and analyse vast datasets far surpasses human capabilities, offering deeper insights that can guide research toward groundbreaking discoveries.
The Reality and the Way Forward:
Stacie Calad-Thomson, business development lead for Nvidia’s Healthcare & Life Sciences Divisions, aptly summarized the sentiment: “AI won’t replace chemists, but chemists who know AI will replace those who don’t.” While this perspective is hopeful, the rapid advancement of AI raises critical questions about how humans will maintain relevance in an increasingly AI-dominated world.
Though definitive answers remain elusive, it is essential to embrace AI as a collaborative tool rather than a threat. By leveraging AI’s strengths to complement human abilities, we can tackle more complex and strategic tasks, accelerating innovation while preserving human creativity and agency. This approach not only enhances productivity but also fosters a dynamic future where AI and human expertise work in concert to achieve remarkable advancements.
Dr. Ashok Kumar FRSC, President of the Centre for Research & Development at Ipca, earned his PhD from CDRI with Dr. Nityanand and post-doctoral experience from Sussex University with Nobel Laureate Prof. Cornforth. His interests include Chemical, Biotechnological & NDDR research and also integrating spirituality, philosophy, and common sense in his professional and personal life.
