The main objective of this research is twofold. First, we aim to investigate the advantages and limitations of different algorithmic strategies, specifically evolutionary computation (genetic algorithms, differential evolution...) and machine learning (neural networks, SVM...), for identifying promising molecules in the context of drug design. This includes evaluating various molecular representations, such as graph-based structures, neural network–learned embeddings, and traditional formats like SMILES. We will also explore hybrid approaches that combine these algorithmic paradigms to enhance performance. Second, we seek to integrate eye-tracking technologies to involve human users in guiding the molecular search process. By capturing visual attention patterns, we aim to steer the exploration of the vast search space toward regions more likely to contain optimal solutions. To support this, we plan to use population-based metaheuristics inspired by swarm intelligence, combined with machine learning (including deep learning when appropriate), to generate novel molecules that satisfy multiple objectives and are simultaneously optimised.