Publicly available biological resources form a complex maze of highly heterogeneous data sources, interconnected by navigational capabilities and applications. Although it offers scientists multiple valuable options, it becomes difficult for them to select the best resources to obtain and exploit their data of interest. When expressing a scientific protocol, they struggle with consolidating the best information about the scientific objects being studied and implementing it in terms of queries against biological resources. In this paper we present the BioNavigation system that allows scientists to express their queries using an ontology representing the conceptual level of scientific classes and labeled relationships. We developed the ESearch algorithm that generates all possible evaluation paths for a given ontological query and also ranks them based on source metadata metrics. BioNavigation thus helps the user visualize the resources at a higher ontological level, build queries gr...