The digitisation of genetic material (DNA, RNA, proteins) allows that biomedia to become computationally recombinant. For some time I’ve been drawn to exploring the transformation of data via ‘machines’ in both the software and wetware sense.
Recently there has been a lot of coverage about the potential for Excel, ubiquitous software through which millions of people manipulate data daily, to have far reaching consequences by users unwittingly introducing errors. The most relevant to my wetware interests was this story about gene data being truncated through misused Excel formulas. Microarray datasets were uploaded to an Excel spreadsheet, but some gene names were incorrectly parsed during the import process. The date format conversions and floating-point format conversions inherent to Excel office software was culpable. Between 30 – 2000 names were effected across the dataset. The conversions were irreversible; the original gene names cannot be recovered. Rather than the narrative of autonomous algorithms wreaking havoc, what transpires here is that receded (or opaque) algorithmic transformations induce error, sans awareness on the part of the human user. This example resonates with the idea of ‘grey media’ advanced by Matthew Fuller and Andrew Goffey in Evil Media
For Wetware Transductions (working title) Microsoft Excel software is misused to produce an animation of a microscopic process – protein folding. The piece uses found footage of a protein folding, converts it to a digital language (ASCII art), and is recombined into an animation by using macro programming affordances of excel.
I’ve used open source Linux software to convert found protein folding renders into ASCII art. For the mock-up of what I hope it to look like (visible on this page) I followed the methodology explained in wonderful detail over on StarioCek’s webpage. The ASCII aesthetic that AALIB achieves is the most desirable (GIF above), but may be unattainable in excel. In the meantime I have been experimenting with extracting raw text ASCII art for inserting into an excel spreadsheet. You can the progress on that front below.
Here is the latest render, this time accomplished with Excel Macros (note: the video is best viewed at vimeo on fullscreen, where the resolution allows the ASCII characters to be visible)
So nicely as the file animates one hurdle remains. The AALIB rendering represents depth well via its shading. Depth is of pretty crucial importance to molecular biology – journals in the field employ stereoscopic illusions in order to convey such information. There are hurdles to shading the characters using the existing format I’ve pursued with the macros. Surmounting them with regex replacements was untenable in terms of animation speed.
A workaround has been reached with straight copy-pasting rather than value replacements. The provisional outcome can be glimpsed in the below version
An excel sheet is chosen to expose the successes and failures that come about through the choice of data representation and various translational machinery (human or otherwise).
Protein folding is chosen because protein folding is a complex computational problem, and a crucial phase in genetic information transfer. The information in a protein only becomes an object with agency once the folding process completes. Proteins are integral to synthetic biology. Folding takes place over a wide range of timescales, from nanoseconds to seconds or longer, depending on the protein.
Found footage is simulatenously a choice of expediency and one laden with conceptual import. Found footage relates directly to the methodologies of appropriation that have accompanied the digitisation of cultural content (the oft called ‘remix culture). It is desirable to evoke this for the proposed artwork as it foregrounds the possibility of recombination once a threshold of ubiquitious digitised objects is reached.