Hundreds of startups are investigating the use of machine learning in the pharmaceutical industry, with many attracting significant investment. By predicting drug behavior, AI has the potential to make drug discovery more efficient and cost-effective, reducing the need for extensive lab work. This article explores the use of AI in drug discovery, from matching patients with precise drugs to designing new molecules, as well as the potential of this new approach.
A breakthrough in precision medicine
Researchers at the Medical University of Vienna have tested a new matchmaking technology developed by Exscientia, a biotech company based in the United Kingdom. They divided a patient’s tissue sample into more than a hundred pieces, which they exposed to various drug cocktails. By using robotic automation and computer vision, they were able to test multiple treatments simultaneously, ultimately identifying an effective drug for the patient.
This AI-assisted patient-drug matching technology is considered a significant breakthrough in cancer treatment, as it allowed for an exhaustive search which ultimately led to a successful remission in one patient.
Overhauling the drug development pipeline
Exscientia intends to change how drugs are created from start to finish. The company is using machine learning for more than just patient matching when it comes to pharmaceuticals; it’s also developing brand-new medicines.
Machine learning is already being incorporated into clinical research systems to speed up the drug discovery and design process, providing a significant increase in potential choices. Since 2021, two drugs developed by Exscientia have been among the first to enter clinical trials after being designed using artificial intelligence.
Future prospects
Despite the fact that AI shows promise in drug discovery, successful clinical research continues to depend on experiments involving cells, tissues, and humans. AI won’t be able to completely replace these procedures, but it does hold several exciting possibilities for how clinical research will be conducted in the future. These include:
Target identification – AI algorithms can help identify promising drug targets by mining large databases of scientific literature and genetic information. Natural language processing techniques enable the extraction of valuable insights from biomedical data, revealing connections and relationships that humans may have missed. This approach can guide researchers in selecting the most promising targets for specific diseases.
Rational drug design – AI can aid in the design of novel drug molecules with specific properties and interactions. The number of possible molecules for drug development far exceeds the current ingredient list used by pharmaceutical companies. Generative AI models can manipulate molecular structures, exploring vast chemical spaces to find molecules that have the desired therapeutic effects. This approach can lead to the creation of more effective and targeted drugs.
Personalized medicine – AI has the potential to tailor drug therapies to individual patients. By analyzing genomic data, patient characteristics, and treatment outcomes, machine learning algorithms can identify patterns and predict the most effective treatments for specific patient populations. This approach can lead to more personalized and precise medicine, maximizing therapeutic benefits and minimizing side effects.
Repurposing existing drugs – AI can help identify alternative uses for existing drugs, accelerating the process of drug repurposing. By analyzing molecular structures and their interactions, machine learning algorithms can predict potential off-target effects and identify new therapeutic applications for known drugs. This approach can reduce the costs and time associated with developing new drugs from scratch.
Enhanced clinical trials – AI can optimize clinical trial design by identifying suitable patient populations, predicting treatment responses, and monitoring patient outcomes. Machine learning algorithms are able to analyze real-time data from clinical trials, identifying potential adverse events and improving patient safety. This approach can lead to more efficient and effective clinical trials, expediting the regulatory approval process.
Drug manufacturing and quality control – AI can improve drug manufacturing processes and ensure quality control. Machine learning algorithms have the ability to optimize production parameters, monitor manufacturing processes in real-time, and detect deviations or anomalies that may affect drug quality. This approach can improve efficiency, reduce costs, and ensure consistent product quality.
Peeking ahead
Although there are still obstacles to overcome regarding data bias, interpretability, and ethical considerations, AI has the potential to significantly improve the pharmaceutical industry by reducing the time, cost, and inefficiency of the drug development process. Exscienta’s exemplary use of AI across the drug development process is leading to the discovery of entirely new classes of medicines that will have a profound impact on people’s health and quality of life. Yet they are not alone.
Insilico Medicine, a drug discovery firm based in Hong Kong, is likewise using AI to develop drugs. Verseon, a California-based drug company, uses generative AI to test millions of potential molecules for drug development. Generate Biomedicines, a startup based in Cambridge, Massachusetts, is also developing drugs with the help of generative AI to design protein structures with specific properties.
Although this is by no means an exhaustive list, it does show that the effective use of AI puts us well on our way to ensuring the efficacy of treatments for cancer and genetic diseases through the development of precision medicine.
