Witjes Vera M, Boleij Annemarie, Halffman Willem
Institute for Science in Society, Faculty of Science, Radboud University, 6500 GL Nijmegen, The Netherlands.
Department of Pathology, Radboud Institute for Molecular Life Science (RIMLS), Radboud University Medical Center, 6500 HB Nijmegen, The Netherlands.
Animals (Basel). 2020 Dec 17;10(12):2415. doi: 10.3390/ani10122415.
Irreproducibility is a well-recognized problem in biomedical animal experimentation. Phenotypic variation in animal models is one of the many challenging causes of irreproducibility. How to deal with phenotypic variation in experimental designs is a topic of debate. Both reducing and embracing variation are highlighted as strategies for reproducibility. In this theoretical review, we use variation in mouse microbiome composition as an example to analyze this ongoing discussion, drawing on both animal research and philosophy of science. We provide a conceptual explanation of reproducibility and analyze how the microbiome affects mouse phenotypes to demonstrate that the role of the microbiome in irreproducibility can be understood in two ways: (i) the microbiome can act as a confounding factor, and (ii) the result may not be generalizable to mice harboring a different microbiome composition. We elucidate that reducing variation minimizes confounding, whereas embracing variation ensures generalizability. These contrasting strategies make dealing with variation in experimental designs extremely complex. Here, we conclude that the most effective strategy depends on the specific research aim and question. The field of biomedical animal experimentation is too broad to identify a single optimal strategy. Thus, dealing with variation should be considered on a case-by-case basis, and awareness amongst researchers is essential.
不可重复性是生物医学动物实验中一个公认的问题。动物模型中的表型变异是导致不可重复性的众多具有挑战性的原因之一。如何在实验设计中应对表型变异是一个有争议的话题。减少变异和接受变异都被视为提高可重复性的策略。在这篇理论综述中，我们以小鼠微生物组组成的变异为例，结合动物研究和科学哲学来分析这场正在进行的讨论。我们对可重复性进行了概念性解释，并分析了微生物组如何影响小鼠表型，以证明微生物组在不可重复性中的作用可以从两个方面来理解：（i）微生物组可作为一个混杂因素；（ii）结果可能无法推广到具有不同微生物组组成的小鼠。我们阐明，减少变异可将混杂因素降至最低，而接受变异可确保可推广性。这些截然不同的策略使得在实验设计中应对变异极其复杂。在此，我们得出结论，最有效的策略取决于具体的研究目的和问题。生物医学动物实验领域过于宽泛，无法确定单一的最佳策略。因此，应根据具体情况考虑应对变异，研究人员的意识至关重要。
