Paper short abstract:

New sequencing technologies change our understanding of cancer and hold the promise of better diagnosis and treatment. However, considerable parts of clinical decisions are delegated to technological means. As a result, drug selection becomes increasingly opaque for the treating physician.

Paper long abstract:

At this time we are observing a fundamental shift in the diagnosis of cancer that holds substantial implications for treatment choices. As swift and comprehensive sequencing technologies become available that change our understanding of cancer. However, extensive research is necessary to produce the body of knowledge that allows classifying DNA-mutations, relating these to phenotypes of cancerous tissue, and ultimately drawing conclusions for treatment.

This research effort has substantial implications. The body of knowledge produced is so vast that the translation into clinical practice cannot be made without computational assistance. Faced with this situation oncologists need the contribution of IT-specialists, among others, who can help to manage and translate the respective data into clinical routine. Bioinformatics is key for identifying evidence of drug-cell interactions that may lead to new treatment options.

Even though this holds promising opportunities for cancer treatment, there are some considerable implications: 1) The search for newly discovered interactions is way too labour intensive for an individual to perform. 2) The required search algorithms can only be performed in a network to which machines belong equally as human beings. 3) As a result, the process of drug selection becomes increasingly opaque for the treating physician. It is the latter who needs to provide a justification for a chosen drug application. Yet, considerable parts of this task become delegated to machinery that requires further specialists trained in bioinformatics and (whole) genome sequencing.

Paper short abstract:

This presentation will look at 'data hugging' practices as a challenge for data driven medicine.

Paper long abstract:

Data-driven medicine has been hailed as the next new frontier in which major new breakthroughs are expected in relation to the development of personalized medicine. An interesting feature in relation to the mobilization of big data is the surprising fragmentation that exists between different types and domains of data on humans, ranging from patient healthcare records to population based biobank cohorts to national data registers on cancer incidence. Recent efforts in a number of countries and regions, such as the EU to mobilize these resources seek to facilitate better access and increased use of such information. Given the fragmentation of these systems, there remains a curious challenge which has been identified already a decade ago relating to samples and information; collections and information are not neutral, but rather come with political, strategic and economic interests which may impact the ways in which information and samples can be used (cf. Hoeyer, 2007).

Drawing on interviews and research conducted as part of the Global Genes, Local Concerns project at the University of Copenhagen, this presentation will examine how the term 'data hugging' has come to represent a particular type of problem in relation to gaining access to samples and data in both large prospective cohorts (LPC), as well as in the Danish health care sector. I will argue that despite policies relating to sharing large data sets and tissue samples across domains to serve as the basis for data driven medicine, there remain numerous hurdles which remain unattended to in relation to recognition of rights and entitlements.

Paper short abstract:

The paper deals with visions, plans and blueprints of data management ‘architectures’ in biomedicine and personalized health. The focus of the analysis is how a ‘national genome server’ has emerged in to architectures and how it shapes the visions of health care.

Paper long abstract:

The prospects of 'personalized medicine' are seen to be embedded in collection and combination of masses of genomic data with many other types of personal health data (patient records with laboratory measurements data, life style surveys etc.) and in circulation and management of such massive data in both research and clinical settings. In this paper I analyze visions, plans and blueprints of data management 'architectures' that are claimed to enable the functioning of future data-driven medicine. The focus of the analysis is a 'national genome server': we analyze how such an imaginary entity appeared in the visions and has been argued for in the prospective landscape of personalized medicine in Finland. The research data consists of 1) graphic charts presented by policymakers and experts when organizing, designing and discussing the 'architecture' of data management in biomedicine and health care, and 2) health and innovation policy discourses which provide a context for these charts. Theoretically, the discussion is based on ANT, Jasanoff's ideas of sociotechnical imaginaries and Helén's idea of experimentality where practices entwine together with scientific, political and economic expectations of biomedicine. Deploying these approaches, the paper addresses the following question: How does the 'national genome server' configure in and effect the network of reasoning and practices around the making of a national system of health data management in Finland? What are the technical and political aspects and functions of the 'national genome server' and what are its implications to health care?

Paper short abstract:

This talk will discuss the process of quality control in a cellular and molecular medicine laboratory. It investigates this process as it emerged in laboratory meetings, an interactional ‘shop floor’ of bioinformation.

Paper long abstract:

This study marks a continuation of my study of post-genomic biology in flight. It captures laboratory members' recognition of a pivotal moment in a project that had gone previously unnoticed. As laboratory members watched a short segment of video from a past meeting, they saw foreshadowed a practical solution to a bioinformatic problem that was implemented months later. By the time a laboratory member implemented the solution, its momentary surfacing in that earlier meeting had been largely forgotten.

By focusing on laboratory meeting, this study proposes to trace the pragmatic, interactional work of translational medicine. It presents a problem of quality control as it was discussed and modulated in one laboratory meeting and follows it to a moment after it was implemented. Digitalization, as an essential element of the work of cellular and molecular medicine, is specified in the prospective and retrospective interactions of the 'shop floor' (Garfinkel, Lynch, and Livingston 1981). That is, bioinformatic work is displayed as it is interactionally constituted and later recognized. In this way, the work of data science in a cellular and molecular medicine laboratory becomes observable, recognizable, and apparent as co-operative action not only by the ethnographer but also by laboratory members.

Paper short abstract:

The relationship between biobanks as bioinformation depositories and narratives that are used to make sense of biobank activities is problematic in Finland. The new ways of organizing medical research challenge the usefulness of the well known narrative of population’s unique genomic heritage

Paper long abstract:

One of the great medical narratives in Finland tells a story about isolated, unique gene-pool of the Finns and its importance to the success of Finnish biomedical research. This genetic heritage has been seen as a special ingredient contributing to the successes and break-throughs in science (i.e. discovering the "Finnish disease heritage"). Today this narrative figures in the rhetorics of biobank actors in Finland, such as people working for biobanks or innovation funds, who seek to legitimize their efforts and convince the possible partners and the public about the usefulness and value of Finnish sample collections for biomedical research.

In the presentation I will discuss how the narrative of unique genome becomes vague when biobank actors situate the value, usability and characteristics of the samples in their collections to the wider field of biomedical research. I suggest that emerging big-data environment of P4-medicine, as well as narrowly defined and very specialized research populations, challenge the strategies the biobank actors and stakeholders promoting biobanks use in their rhetorics. Their narratives underline national strengths of Finland to promote the usage of bioinformation from Finnish clinical biobanks.

I will ask how national uniqueness matters in the digitalized environment where bioinformation might be pooled from various sites together to form the "research population". Thus the presentation contributes to the STS discussions on research populations and data. The presentation is based on various kinds of data; published materials, fieldnotes from the participant observation as well as interviews.