Presently, it is widely accepted complex systems couldn't be comprehended by studying parts in isolation without examining integrative and emergent properties, and system-level understanding thus has become the focus in biological science. However, it should also be noted that common systematic analysis was restricted to large-scale analysis at a certain level, while the facts that the nature of complex systems is their multi-scale structures was usually neglected or ignored. Therefore, this paper described a multi-scale methodology to investigate the nature of biological complexity and prospected this methodology could lead to a promising revolution in current system-level understanding and the integration of molecular biology databases.