We present a provenance model based on ontology representation of execution records of in silico experiments. The ontologies describe the execution of experiments, and the semantics of data and computations used during the execution. This model enables query construction in an end-user oriented manner, i.e. by using terms of the scientific domain familiar to researchers, instead of complex query languages. The presented approach is evaluated on a case study Drug Resistance Application exploited in the ViroLab virtual laboratory. We analyze the query capabilities of the presented provenance model. We also describe the process of ontology-based query construction and evaluation.