SHAPING NOISE
​
John Reinhard
Saccharomyces cerevisiae
?
​
omics
?
?
Saccharomyces cerevisiae
​
- WARNING: This video may potentially trigger seizures for people with photosensitive epilepsy. Viewer discretion is advised.
​
- We recommend the use of headphones while listening to this video)
Archeo-nomics
Preface​
​
In the past few decades in the study of life sciences, new experimental practices have enabled the rapid collection of unprecedented amounts of biological data. Countless terabytes of high-resolution videos of microscopic cells are stored as they wiggle and interact; millions of genomes have been sequenced; and thousands of strains of bacterial DNA have been reengineered and recorded in action. These vast, immense data sets constitute new archives of knowledge where, as scientists, we expect to find answers to fundamental questions of life.
Often, as it happens, only fractions of these data and images are relevant and meaningful to researchers. Computerized analysis is the only way to make sense of the data and distinguish relevant from irrelevant data. The immensity and unintelligibility of these archives to human eyes resurfaces in a sharp key the question of signal to noise ratio. While extracting information from the environment that surrounds us, how do we learn to distinct signal from noise, according to which definition of signal? According to which description of environment? and how much are we taking part in constructing the context?
One may as well reflect in front of such flabbergasting capacities, upon how human collective practices in hunting down signals in various contexts have evolved across the centuries and how do they, if at all, reflect upon individual access to knowledge.
​
This video was put together from microscopic images that were generated in a systematic high-throughput experiment. These experiments aim to discover previously unknown roles of certain genes or gene products in cellular structures and functions. During this experiment roughly 50.000 images of different baker's yeast (Saccharomyces cerevisiae) strains were generated, analyzing every single gene of this organism at the same time.
​
​
Similar experimental designs, e.g. so-called omics approaches, produce large amounts of data. In this act, which I would call a reflective performative process, I worked with about 3.000 discarded images of the aforementioned yeast experiment where either no cells were captured, by the automated microscope, or something else distorted the image.
​
The archive analogy
​
These datasets are collections of information that, in a way, resemble archives.
Furthermore, they are part of scientific publications, that are collected in databases that are built and maintained by many different parts of society. I went into one of these archives to mine material, take it out of context and let it resurface in another light. By changing the context, a different life had been bestowed upon these data-corpses, treating them as plastic material that can be molded.
Measurements and the human impact
​
Visible features which can be recognized while looking at the microscopic images composing the video would be considered artifacts. That is in the natural sciences definition: an error introduced by the equipment or technique being utilized. Interestingly, the archaeological definition of artefact is profoundly different: something that is made by humans.
Reviewing these experiments or datasets as our current human approach to get to know something about the world, it is apparent that they contain lots of noise. The researcher is interested in the significant signals within that noise and thus it is her performative act to extract these signals from the noise. We do then find ourselves suspended between the archeological and the scientific notion of artefact, while in the production of new knowledge.
I wish to thank Maya Schuldiner for letting us use the discarded microscopic images that were generated in the experiment conducted by her lab.
