Archimedes optimization algorithm for DNA motif discovery

This research work introduces a novel metaheuristic method named Archimedes Optimization Algorithm (AOA) for solving biomedical problems. Two case studies on Motif Discovery (MD) and Optic Disc (OD) are also analysed. AOA is based on the concept of Archimedes’ principle in physics. It assumes that a buoyant force exerted upwards on an object in a fluid is equal to the weight of the displaced fluid. In our case studies, the motif part encapsulates the motifs’ essential features in DNA sequences, the suggested solution emulates the huddling behaviour of Archimedes’ law while incorporating the motif characteristics to detect the desired motif. The effectiveness of the AOA optimizer is tested using both real & synthetic datasets of motif discovery, and optic disc. Based on the results comparison with various common algorithms (including DREME, PMbPSO, MEME, MACS, and XXmotif), it can be confirmed that AOA can efficiently solve hard motif discovery problems. In addition, efficiency and flexibility of the proposed AOA are further validated on a biomedical problem, Optic Disc (OD) extraction from 369 retinal fundus images of four publicly available datasets. Results confirm the robustness of AOA and demonstrate its value as a OD finding tool with a 100% success rate for images from well-known datasets like DRIVE, and DRIONS-DB datasets, while achieving 98.5% and 98.8% for DIARETDB0 and DIARETDB1 datasets respectively; which is surpassing the state-of-the-art methods.