Machine learning for genomic profiling and drug discovery in personalised lung cancer therapeutics.

Abstract

Lung cancer is a leading cause of cancer-related mortality, with approximately 2 million new cases and 1.8 million deaths annually, and studies suggest that by 2050, these numbers will reach 3.8 million cases and 3.2 million deaths. The high mortality rate highlights the urgent need for early diagnosis and rapid drug development. Genomic approaches provide insights into tumour biology, supporting personalised medicine. This study explores the role of machine learning (ML) in enhancing genomic analysis and drug discovery for lung cancer treatment. A comprehensive PubMed search was conducted to identify relevant publications from the last 10 years. Selected studies were critically reviewed to understand how ML algorithms are applied in lung cancer genomics and drug discovery. ML algorithms such as random forests, gradient boosting, support vector machines, autoencoders, CNNs, and RNNs are widely used for genomic pattern identification. Techniques like reinforcement learning, deep neural networks, GANs, and GNNs are employed for drug discovery. ML models have achieved over 95% accuracy in certain lung cancer applications. However, challenges remain, including data scarcity and model interpretability. ML significantly enhances lung cancer's genomic analysis and drug design; however, further optimisation and clinical validation are essential for effective real-world implementation.