Multivariate statistics and the enactment of metabolic complexity

Abstract

This ethnographic study, based on fieldwork at the Computational and Systems Medicine laboratory at Imperial College London, shows how researchers in the field of metabolomicsthe post-genomic study of the molecules and processes that make up metabolism -enact and coproduce complex views of biology with multivariate statistics. From this data-driven science, metabolism emerges as a multiple, informational and statistical object, which is both produced by and also necessitates particular forms of data production and analysis. Multivariate statistics emerge as 'natural' and 'correct' ways of engaging with a metabolism that is made up of many variables. In this sense, multivariate statistics allow researchers to engage with and conceptualize metabolism, and also disease and processes of life, as complex entities. Consequently, this article builds on studies of scientific practice and visualization to examine data as material objects rather than black-boxed representations. Data practices are not merely the technological components of experimentation, but are simultaneously technologies and methods and are intertwined with ways of seeing and enacting the biological world. Ultimately, this article questions the increasing invocation and role of complexity within biology, suggesting that discourses of complexity are often imbued with reductionist and determinist ways of thinking about biology, as scientists engage with complexity in calculated and controlled, but also limited, ways.

Rethinking Interdisciplinarity across the Social Sciences and Neurosciences

“…Our hunch is that the sense of possibility that we witnessed has moved to other, currently 'sexier' , biosocial fields. We think here, for example, of epigenetics (Landecker and Panofsky 2013) or metabolomics (Levin 2014). It is noteworthy that neither the ENSN nor the European Platform exists any longer (although a group of junior researchers, based primarily at King's College London, is attempting to resuscitate a 'Neuroscience and Society Network' -see Mahfoud and Maclean 2015).…”

Making the microbiome public: Participatory experiments with DNA sequencing in domestic kitchens

“…In conceiving our methodology, we were influenced by three overlapping literatures within the burgeoning research and practice on public engagement with science and technology. The first is a long‐standing interest in science and technology studies and its focus on opening up laboratory spaces and the black boxes of scientific technologies, like those involved with DNA sequencing (Levin, ; Mackenzie, ; Rabinow, , ). There is a long history of ethnographic work in which social scientists shadow laboratory scientists to better understand the social norms, material assemblages and political interests that come to frame the production, circulation and governance of science (Latour & Woolgar, ).…”

Ethnographic Agency in a Data Driven World

“…For my PhD, I did an ethnography of how scientists worked with large datasets and statistics. I began my research at a laboratory at Imperial College London thinking that I would study academic-industry collaborations, and ended up focusing on how scientists were trying to understand the complex system of metabolism with multivariate statistics (Levin 2014a). I carried out participant observation with scientists in the field of "metabolomics," as they attempted to understand the role that metabolism played in enabling living beings to interact with their environment over time.…”

“…I pored over scientific literature and attended training courses and seminars. Even though I had no formal training in statistics, I learned to "speak the language of data" by familiarizing myself with the theory behind principal components analysis, supervised learning techniques, and neural networks (Levin 2014a). By gaining a deeper understanding of virtual and intangible data-rich systems, I was able to reflect on how quantitative data was reshaping concepts like metabolism and health, creating friction between scientists and clinicians (Levin 2014b), and shaping notions of "persons" and "populations" in healthcare systems.…”

“…This distance from data is further evidenced in the anthropological and social studies of science (STS) literature on quantitative data, which has tended to focus on its end products-papers and visualizations, societal implications (Miller 2015;Dougherty 2015)instead of the everyday practices that go into negotiating and making sense of quantitative data (Levin 2014a;Starosielski 2015;Buur, Mosleh, and FYHN 2018). This has left ethnographers, in both industry and academic settings, with few playbooks for engaging critically with quantitative data, or for doing "participant observation" on the practices that shape how quantitative data are made and used.…”