The recent popularity of systems approaches in molecular biology is perhaps best understood as a reaction to technological developments beginning in the 1990s, notably the large sequencing projects such as the Human Genome Project. Certainly, this challenge requires new analytical tools to translate 'big data' into biological understanding, but it has prompted calls for more fundamental change at the conceptual level as well. Systems biology is often presented as an alternative and superior way of doing biology.

The term 'systems biology' appeared towards the end of the 1990s and gained widespread use in the early 2000s. Very early on, systems biology showed the characteristic features of an institutionalized discipline: research institutes for systems biology were founded, conferences about systems biology began to be held, and several journals specifically dedicated to systems biology were created. However, in spite of this concretization at the institutional level, no clear and unique characterization has crystallized up to now.

The field shows a considerable heterogeneity of approaches that have their historical roots in different traditions of theoretical biology and of other theoretical fields studying complex systems. To be sure, there are a number of distinctive features that are commonly cited, such as the combination of mathematical methods with experimental approaches, the investigation of quantitative and dynamic properties of living systems, and a focus on interdisciplinarity and integration.

However, aside from these very general attributes, systems biologists often highlight very different aspects as being central to the new field. Some see the main goal of systems biology in the integration of different levels of biological information; while others emphasize the continuity with predecessors of systems biology, such as the 'general systems theory' developed by the Austrian biologist Karl Ludwig von Bertalanffy; still others stress the importance of engineering concepts, like robustness, modularity, or feedback. Independently of a precise definition, however, one might ask whether systems biology really provides a radically new perspective on living systems and, if so, how it relates to the 'traditional' approach of molecular biology. Systems biologists often invoke the opposition between 'reductionism' and 'holism' to argue for the difference and superiority of their approach. According to the consensus that has emerged from this kind of rhetorics, traditional molecular biology has confined itself to the study of parts, whereas systems biology aims at understanding how those parts interact to produce phenotypic behavior. In this way one can conveniently equate the two labels of 'molecular biology' and 'systems biology' with competing philosophical perspectives. A closer look at how systems biology is actually pursued in practice, however, reveals that the dichotomy is oversimplified, overstating superficial differences while blurring relevant ones. Molecular biologists have never been interested only in a taxonomy of parts, instead their motivation has been to explain behavior by referring to molecular mechanisms. Therefore, molecular biologists and systems biologists share important epistemic goals. Relevant differences, on the other hand, arise around the question of how to best discover the underlying mechanisms that account for these phenomena. Molecular biologists have pursued the search for mechanisms by making certain assumptions about biological organization. In particular, they have assumed that living systems can be studied by conceptually dividing them into a set of small and quasi-independent mechanisms. Furthermore, describing biological processes with the metaphor of linear chains of information processing, licenses a focus on the qualitative features of these mechanisms. It is important to highlight that such assumptions, even though unwarranted in some contexts, are precisely what makes scientific discovery efficient and successful.

The crucial and often neglected point is that alternative approaches, such as the ones classified as 'systems biology,' must introduce many simplifying assumptions as well in order to be efficient. Interestingly, different approaches within systems biology, especially those dealing with large biological networks, often make diametrically opposed simplifying assumptions about the organization of living systems at the molecular level. Systems biology cannot do without simplifying assumptions and in this respect is not different from traditional molecular biology. But perhaps one of the most important merits of systems biology so far has been to draw attention to alternative ways of reducing the complexity of biological discovery. Furthermore, systems biology provides a great opportunity of unifying biological knowledge by creating formal models that make underlying assumptions explicit and facilitate the integration of different approaches.