Blog Post

The potential of generative artificial intelligence (AI) to create language and artwork that resembles that of a human has recently drawn interest from all over the world. Beyond these widely used uses, artificial intelligence (AI) is transforming drug research and could change the process of creating life-saving drugs.

AI-enabled drug discovery is a reality today, not just a pipe dream. In this field, a number of historic firsts have been accomplished in recent years. Exscientia reported in the beginning of 2020 that the first therapeutic molecule created by AI had entered human clinical trials. The next year, 330,000 protein structures, including all 20,000 found in the human genome, were predicted by DeepMind’s AlphaFold algorithm. Since then, the system has grown its database to include almost all cataloged proteins that are known to science.

Additional progress was made in 2022 when Insilico Medicine started Phase I clinical trials for the first drug found by AI, based on an AI-discovered novel target. This was completed more faster and for a much lower cost than conventional preclinical programs. In early 2023, AbSci became the pioneer in developing and validating de novo antibodies in silico using generative AI. A short while later, a medication developed and found with AI received the first-ever Orphan Drug Designation from the FDA, and Insilico Medicine intends to launch a global Phase II study for the medication shortly.

These accomplishments are not one-off events. In March 2022, there were over 150 small-molecule medications in discovery and over 15 in clinical trials among biotech companies that prioritize artificial intelligence. This quick development highlights how revolutionary AI can be for drug discovery.

Conventional drug discovery is infamous for being costly and time-consuming; pre-clinical phases can cost anywhere from hundreds of millions to billions of dollars and persist for three to six years. But artificial intelligence (AI) tools are transforming almost every step of this process, with the potential to drastically alter the pharmaceutical industry’s economics and speed.

AI is being used in drug discovery at different phases. AI algorithms are trained on large datasets for target discovery in order to comprehend disease causes and find new proteins or genes that can be addressed. More advanced systems, such as AlphaFold, anticipate targets’ three-dimensional structures, which speeds up the development of suitable medications. AI also enables high-fidelity molecular simulations, minimizing the requirement for physical testing of prospective therapeutic molecules.

Without the requirement for simulated testing, several AI systems are able to anticipate important aspects of drug candidates, such as toxicity and bioactivity. AI can also be used to create brand-new, previously undiscovered medicinal compounds from scratch, which changes the focus from conventional chemical library screening. AI outperforms earlier ranking algorithms in helping to rate and prioritize attractive lead compounds for additional evaluation once they have been identified. AI is also being used to create synthesis routes for the production of fictitious medicinal molecules, occasionally recommending adjustments to make them simpler to produce.

The idea of completely automated end-to-end drug research is getting closer to reality as AI systems advance. Investors have taken note of this possibility. Morgan Stanley estimates that over a ten-year period, even small increases in the success rates of early-stage drug development made possible by AI might lead to the discovery of 50 more innovative medicines, so creating a potential market opportunity valued at over $50 billion. The investment environment reflects this zeal, with third-party funding for AI-enabled drug research more than doubling yearly for five years running, to reach over $5.2 billion by the end of 2021.

Many businesses have raised hundreds of millions of dollars to develop their AI-driven drug discovery pipelines, including XtalPi, AbCellera, Relay Therapeutics, Atomwise, Recursion Pharmaceuticals, Schrödinger, and ExScientia. This funding inflow demonstrates the increasing belief in AI’s ability to revolutionize the pharmaceutical sector.

AI’s contribution to drug discovery has far-reaching ramifications. If current trends continue, it’s conceivable that many of the medications we take will be machine-designed rather than human-designed in the near future. This change may result in more accessible medications and the ability to treat diseases that are currently incurable by lowering costs and shortening development times.

But there are drawbacks to this technical advancement as well. Intellectual property rights, the possibility of technology abuse, and guaranteeing the continuous safety and effectiveness of pharmaceuticals in this new paradigm are unresolved challenges. As we approach this new chapter in the history of drug discovery, it is imperative that advocates, legislators, and other relevant parties start getting ready to tackle these issues right away.

AI-enabled drug discovery has a bright future ahead of it; in fact, it is now here. Now, the question is whether we’ll be prepared to take advantage of the benefits it offers and successfully manage the hazards that go along with it. Our capacity to successfully negotiate this new terrain will have a significant impact on pharmaceutical companies, medical innovation as a whole, and global health. It is obvious that as time goes on, the nexus between artificial intelligence and drug discovery will require a great deal of attention, resources, and thought in order to fully realize its promise for improving human health.

Reference: https://blog.petrieflom.law.harvard.edu/2023/03/20/how-artificial-intelligence-is-revolutionizing-drug-discovery/