System-driven Research: Legitimate Experimental Design for Biological/biomedical Research
MetadataShow full item record
Exciting new high-throughput methodologies permit innovative experimental designs for current biological and biomedical research. Much of this research is “non-hypothesis-driven”: its experiments are not designed around a specific testable hypothesis. This challenge to the hegemony of the hypothesis poses important problems. For philosophers, it yet again questions the hypothetico-deductive method (HDM), the established hypothesis-driven strategy for science. For scientists, it raises practical issues about how to evaluate non-hypothesis-driven experimental design and its data, including what research deserves funding. Overall, it asks what qualifies as worthy science. I define hypothesis-driven research as research where a well-articulated hypothesis immediately governs the experimental design. Several different styles of non-hypothesis-driven research exist. With system-driven research, the biological entity under investigation, demarcated as a system, governs the experimental design. My thesis is that system-driven research, which typically involves high-throughput methodologies collectively known as “omics”, constitutes scientific research as worthy as hypothesis-driven research. The HDM derives its cogency from its logical structure. I argue that system-driven research is supported by a similarly powerful scientific epistemology, the Omics Experimental Strategy (OES). I aim to show how the OES operates. I anchor this analysis in a subdiscipline of proteomics, metalloproteomics, which I helped to develop in the early 2000s. Biological research attempting to examine a complex biological system directly reflects a distinctive epistemological orientation. The OES attempts to be fully inclusive. It proceeds without prejudice of expectation related to a hypothesis and may reveal ‘surprising’ findings, otherwise inaccessible to a chain of hypotheses. With the OES, excellent experimental design, technical adequacy and adherence to standards of expert performance in the field are crucial for ensuring that data are reliable. Pattern-detection is central to data evaluation. For knowledge production, contextualization of the findings in the system itself is critical. With the HDM, the findings are evaluated in terms of the hypothesis; with the OES, the findings are evaluated in the context of the system. System-driven research constitutes proper scientific research. It operates alongside hypothesis-driven research, not merely preparatory to it. Importantly, system-driven research offers innovative, unique ways to understand biological systems, and very possibly the methodology of science itself.