Key element identification in large biological datasets: An MCDM comparative study

Modeling biological systems as networks enables efficient analysis, leveraging graph theory. Centrality measures identify key elements, but combining multiple metrics is necessary for comprehensive understanding. This research utilizes multicriteria decision making (MCDM) to rank key proteins in the Escherichia coli network, considering four key network metrics as multi-attributes: the frequency of direct connections (degree centrality), the extent of intermediary roles (betweenness centrality), the proximity to other nodes (closeness centrality), and the influence based on connections to high-scoring nodes (eigenvector centrality). The main objective is to compare two established MCDM methods, TOPSIS and VIKOR, investigating their differences and similarities in identifying and ranking key proteins Results reveals that both methods exhibit similar performance, with VIKOR showing a slight edge in most cases. Notably, VIKOR's superior performance in identifying top-ranked key proteins suggests its potential for applications in biological networks.