Mobile Menu

AI in Drug Discovery 2024: Where are we now?

Back in 2020, we covered an industry analysis of the role of Artificial Intelligence in Drug Discovery. The report covered the period up to 2024 and suggested that the market was worth hundreds of millions of dollars.

What’s going on in this field now that we’ve reached 2024? In this feature, we look at the emergence of AI as a tool in the drug discovery process and take a look into the future.

How can AI be used in drug discovery?

The traditional drug discovery journey is a tough one. Most drugs take anywhere from 10-15 years to make it to the clinic, cost millions (if not billions) of dollars to create and, ultimately, most candidates fail due to safety concerns or efficacy issues. The integration of genetic data has high potential to improve the process – a 2019 study suggested that candidates supported by genetic data were twice as likely to make it to the clinic – but the journey from bench to bedside is still long and arduous.

Enter artificial intelligence. AI is proving to be a powerful tool in the drug discovery process, offering innovative solutions to longstanding challenges. AI algorithms can sift through vast repositories of biological data to identify potential drug targets with unprecedented speed and high accuracy. By analysing everything from genomic to clinical data, AI tools can pinpoint molecules or biological pathways that play a key role in disease progression, providing researchers with valuable insights into potential therapeutic interventions. And that’s not all – AI-guided screening techniques can rapidly scan large libraries of compounds to identify those with the highest likelihood of binding to the target. This process, which was once time-consuming and costly, can now be completed in a fraction of the time, significantly accelerating the pace of drug discovery. Moreover, AI-driven predictive modelling enables researchers to assess the efficacy and safety of drug candidates more accurately, guiding decisions on which compounds to prioritize for further testing.

How are tech giants getting involved?

As the potential of AI tools has grown, so has the number of companies investing in AI solutions and labs using them. For comprehensive analyses of the market, check out some of these reports, which predict that the industry will be worth billions of dollars in the next few years.

Unsurprisingly, big names in the tech world have started to dip their toe in the healthcare sphere, leveraging their technological prowess to make their mark on the drug discovery world. Among these key players are NVIDIA, AWS and even Microsoft. But perhaps the most famous company investing in drug discovery and molecular biology is Google. The tech giant’s AlphaFold2 tool has been a gamechanger for protein structure prediction, which in turn can assist the drug discovery process via the screening of existing compounds and guiding the design of new ones. Additionally, AlphaMissense, which is based on AlphaFold2’s technology, can predict the impact of genetic variants. This could significantly transform the target identification step of the drug discovery process.

Biggest developments so far

What are the biggest developments in AI usage we’ve seen so far in the drug discovery realm? The potential of the technology was starkly highlighted during the COVID-19 pandemic, when AI was used for drug repurposing to try and find a treatment for the deadly virus. The need for swift and efficient screening tools carried over into a number of different fields, including, for example, the repurposing of antipsychotic drugs to treat cancer.

Additionally, researchers are starting to explore new ways to utilise AI in the screening process. This study from July 2023 details the use of deep learning models to sort through thousands of images from phenotypic-based drug discovery processes, as a means to identify cellular changes.

If you want to read more recent reviews about the use of AI in drug discovery, particularly when using genetic data, check out our recent ‘State Of: Genomics in Drug Discovery’ feature.

Perhaps the area with the most potential is the generation of completely new drugs designed by AI. A number of companies have explored this concept, and some of these drugs have made it into early clinical trials. But what were the results of these experimental techniques?

What does the future hold?

Ultimately, despite the investment of key players from both the tech and healthcare industries, as of 2024, none of these AI-generated drugs have successfully made it to the clinic. Despite these setbacks, scientists are still fighting to push forward the use of AI in drug discovery, in 2024 and beyond. Insilico Medicine’s AI-generated anti-fibrotic drug was the first of its kind to be administered in humans in a Phase 2 trial last year, and in March 2024, a study in Nature Biotechnology detailed the journey from generation with an LLM to trials.

In 2023, the US Food & Drug Administration published guidelines related to the use of AI in drug discovery. The report highlighted key areas of potential whilst also considering risk mitigation and a need for clearer guidelines related to the use of this technology. With better clarity, researchers should be able to put the technology to better use.

Another key aspect in ensuring best practice in the use of AI is through building trust in models. In a recent Front Line Genomics webinar, Richard Lewis (Director, Data Science, Computer-Aided Drug Design, Novartis), discussed the sociological considerations of model building and AI’s place in drug discovery. If researchers are armed with better understanding and trust in the technology, AI should be able to be used more effectively in the process.

It is clear that in 2024, the use of AI in drug discovery has expanded, but has not yet reached its full potential. Technological and ethical challenges remain, but experts are on their way to solving these problems, and investment in the field is continuing to grow in response.

More on these topics

AI / Drug Discovery / Generative AI