The need for evaluating scientific research demands adjusting methodology. Today, the two existing approaches, namely bibliometrical analysis and expert opinion, will not meet the demands of researchers and administrators if used separately. Based on the reviews, reports and personal practical experience, the author examines the pros and cons of these two approaches and concludes on the need to integrate them so they complement one another. The bibliometrical studies related to scholarly publications processing and quantitative analysis enable to identify the most globally important works while the expert opinion enables to distinguish radically new themes at the early stage. The author argues that the methodology of three components will be the most efficient for science advance; they are: bibliometrical analysis, expert review, and opinion of the professional who would browse the publications in his/her specialty searching for promising themes. Properly synchronized, these three approaches would enable to make economically efficient and informative foundation for rating research activities and to identify the newest vectors in science at the early stage of their emergence.