Artificial intelligence (AI) has recently shifted from a futuristic concept to an essential tool across industries, particularly in drug discovery. Last year’s Nobel Prize in Physics, awarded to John J. Hopfield and Geoffrey E. Hinton for their foundational contributions to machine learning and artificial neural networks [1], highlights the transformative impact of AI in understanding and solving complex scientific problems across diverse fields. Similarly, AI empowers researchers in drug discovery to analyze and predict molecular interactions with extraordinary precision, accelerating the development of new treatments [2-9].
On a global scale, AI is revolutionizing each stage of the drug discovery pipeline (Table 1), transforming vast datasets into life-saving medications.Developing new drugs is a lengthy, expensive, and complex process that often spans more than a decade and costs billions of dollars. A major challenge in this field is the "failure cascade," where potential drug candidates fail at different stages, leading to significant sunk costs. This is where AI is stepping in to revolutionize the process. Leveraging AI allows sifting through massive biological datasets, predicting molecular interactions, and even simulating drug efficacy in silico before moving to the lab. These AI-driven advancements are not only expediting drug discovery but also enabling a more targeted and efficient approach to finding new therapies.
Table 1. AI in Different Stages of the Drug Discovery Pipeline
|
Pipeline stage |
Description and challenge |
Why AI matters |
|
Target Identification |
Identifying the biological targets (e.g., proteins, genes) associated with a disease. Traditional methods for target identification are time-consuming and subject to errors due to the complexities of biological systems. AI’s ability to analyze complex datasets, such as gene expression profiles and protein structures, makes it uniquely suited for this stage. |
Machine learning (ML) algorithms process complex biological data, identifying connections and predicting which proteins or genes could serve as viable targets. For example, deep learning models that analyze protein structures have become indispensable for understanding the dynamics of disease-relevant proteins. Recent studies [10] have shown how AI-driven models predict disease-related targets with greater accuracy, guiding researchers to focus on the most promising leads. |
|
Hit Identification and Screening |
Finding compounds that interact with identified potential targets. Traditional screening involves testing millions of compounds, which is labor-intensive and costly. AI streamlines this process through predictive modeling, where algorithms evaluate compound-target interactions. |
AI models, such as generative algorithms, can predict molecular interactions with impressive precision. This reduces the dependency on exhaustive physical screenings and quickly narrows down the most promising candidates. Generative AI models trained on vast chemical datasets can design novel compounds, predicting their efficacy and selectivity toward specific targets, thus accelerating the hit identification process. By simulating how these compounds interact with targets, drug effectiveness can be estimated before entering lab tests. |
|
Lead Optimization |
Refining the selected compounds to improve their efficacy, selectivity, and safety. This step is critical for creating a viable drug candidate but traditionally requires labor-intensive chemical modifications and testing. |
By analyzing the structure-activity relationship (SAR), AI models enable us to evaluate the impact of specific molecular structure modifications on improvement of a compound’s drug-like properties. This computational approach reduces the trial-and-error nature of traditional optimization, saving both time and resources. AI-driven platforms [8,9] enable chemists to optimize compounds systematically, targeting improvements in stability, potency, and biocompatibility. |
|
Preclinical and Clinical Testing |
Ensuring the drug’s safety, efficacy, and stability in biological systems, which are necessary before approval. AI can support these stages by predicting a compound's toxicological and pharmacokinetic properties, helping to identify potential risks early. |
In preclinical testing, AI models can predict potential adverse effects and the pharmacokinetics of compounds, such as absorption, distribution, metabolism, and excretion (ADME) properties. In clinical testing, AI is used to optimize patient recruitment, analyze patient data, and help design trials that are more likely to succeed. AI algorithms have been shown [11] to analyze patient data to identify those most likely to respond to the drug, allowing for more targeted and efficient trials. This not only increases the chances of success but also reduces the time and cost involved in testing. |
AI's transformative impact on drug discovery is evident across all drug discovery stages, from target identification to clinical testing. Looking ahead, the future of AI in drug discovery holds exciting possibilities, driven by advancements in algorithms and integration with other cutting-edge technologies.
To open up and develop even more prospects, Life Chemicals offers AI-enhanced screening libraries, specifically designed in collaboration with Variational AI, to include target-selective compounds. These libraries harness AI to predict target interactions accurately, streamlining the initial screening process. In particular, our Novel Generative AI Designed Target Selective Compounds target key kinases and GPCRs of critical interest.
Life Chemicals’ EGFR AI-enhanced Screening Compound Library, for instance, exploits AI to selectively generate compounds that bind effectively to Epidermal Growth Factor Receptor (EGFR), a protein target often implicated in cancer. In particular, AI algorithms aid in designing compounds that inhibit EGFR while overcoming resistance to third-generation EGFR inhibitors [12,13]. AI models analyze molecular patterns associated with resistance mutations and provide guidance to creating next-generation EGFR inhibitors with enhanced efficacy, thus facilitating breakthroughs in targeted cancer treatments (Fig. 1).
Learn More About Our AI Capabilities
Implementing its mission of supporting interdisciplinary pioneering research, Life Chemicals is always ready to facilitate successful solutions for complex drug discovery problems. In this context, the full potential of our AI-driven libraries and their applications is available to address your challenging projects.
Please contact our team at marketing@lifechemicals.com for detailed information and collaboration opportunities.
|
|
|
Fig 1. Representative stock and tangible screening compounds from the EGFR AI-based Screening Library.
References
- Press release: NobelPrize.org. Nobel Prize Outreach AB 2024. Wed. 11 Dec 2024. nobelprize.org/prizes/physics/2024/press-release/
- Deng, J., Yang, Z., Ojima, I., Samaras, D., & Wang, F. (2022). Artificial intelligence in drug discovery: applications and techniques. Briefings in Bioinformatics, 23(1):bbab430. DOI: 10.1093/bib/bbab430
- Jiménez-Luna, J., Grisoni, F., Weskamp, N., & Schneider, G. (2021). Artificial intelligence in drug discovery: recent advances and future perspectives. Expert opinion on drug discovery, 16(9):949-959. DOI: 10.1080/17460441.2021.1909567
- Mak, K. K., Wong, Y. H., & Pichika, M. R. (2024). Artificial intelligence in drug discovery and development. Drug discovery and evaluation: safety and pharmacokinetic assays, 1461-1498. DOI: 10.1007/978-3-031-35529-5_92
- Sellwood, M. A., Ahmed, M., Segler, M. H., & Brown, N. (2018). Artificial intelligence in drug discovery. Future medicinal chemistry, 10(17):2025-2028. DOI: 10.4155/fmc-2018-0212
- Chen, W., Liu, X., Zhang, S., & Chen, S. (2023). Artificial intelligence for drug discovery: Resources, methods, and applications. Molecular Therapy-Nucleic Acids, 31:691-702. DOI: 10.1016/j.omtn.2023.02.019
- Gupta, R., Srivastava, D., Sahu, M., Tiwari, S., Ambasta, R. K., & Kumar, P. (2021). Artificial intelligence to deep learning: machine intelligence approach for drug discovery. Molecular diversity, 25:1315-1360. DOI: 10.1007/s11030-021-10217-3
- Rehman, A. U., Li, M., Wu, B., Ali, Y., Rasheed, S., Shaheen, S., et al. (2024). Role of Artificial Intelligence in Revolutionizing Drug Discovery. Fundamental Research. DOI: 10.1016/j.fmre.2024.04.021
- Gangwal, A. and Lavecchia, A. (2024). Unleashing the power of generative AI in drug discovery, Drug Discovery Today, 29(6):103992. DOI: 10.1016/j.drudis.2024.103992
- Arnold, C. (2023). Inside the nascent industry of AI-designed drugs. Nat. Med. 29:1292–1295. DOI: 10.1038/s41591-023-02361-0
- Serrano, D. R., Luciano, F. C., Anaya, B. J., Ongoren, B., Kara, A., Molina, G., et al. (2024). Artificial intelligence (AI) applications in drug discovery and drug delivery: revolutionizing personalized medicine. Pharmaceutics, 16(10): 1328. DOI: 10.3390/pharmaceutics16101328
- Singh, D., Attri, B.K., Gill, R.K., Bariwal, J. (2016). Review on EGFR Inhibitors: Critical Updates. Mini Rev. Med. Chem. 16(14):1134-66. DOI: 10.2174/1389557516666160321114917.
- Shi, K., Wang, G., Pei, J., Zhang, J., Wang, J., Ouyang, L., Wang, Y., Li, W. (2022). Emerging strategies to overcome resistance to third-generation EGFR inhibitors. J. Hematol. Oncol. 15(1):94. DOI: 10.1186/s13045-022-01311-6
Comments ()