Toward personalised cancer care: AI in precision and molecular oncology

The development of artificial intelligence (AI) in molecular oncology has introduced a powerful new approach to understanding tumour biology. By enabling comprehensive analysis of molecular expression patterns, AI facilitates more precise therapeutic selection, improved personalised therapies, more accurate prediction of treatment response, and reduced toxicity. Central to this capability is the use of large, complex datasets, including multi-omics data, cheminformatics, and clinical trial information, which allow AI models to predict molecular interactions with high accuracy, optimise pharmacokinetic and pharmacodynamic profiles, and inform the rational design of therapeutic compounds. From a methodological perspective, foundational models trained on large and diverse datasets can be adapted and implemented across a wide range of specialised oncology and research tasks. Large language models enhance this framework by interpreting unstructured data from sources, such as electronic health records, radiology reports, and scientific literature, enabling clinicians to interact more intuitively with complex information. In parallel, multimodal algorithms integrate histopathology, imaging, molecular datasets, and clinical records into unified analytical systems, allowing tumour biology to be studied with greater depth and contextual understanding than any single modality can provide. These advances extend beyond data integration to biological insight. AI models have demonstrated the ability to identify key biological nodes within tumours that are critical for growth and survival, including specific molecular pathways, thereby highlighting potential targets for therapeutic intervention. By translating existing biological knowledge into clinically actionable insights, AI supports the development of more tailored and effective treatment strategies. Although these approaches are still evolving, they show significant potential to improve diagnostic accuracy, uncover novel therapeutic targets, and advance the broader understanding of cancer biology, ultimately contributing to more effective and personalised oncology care.