Drawing with DNA: 'Bioart' illuminates genomics

[Guest guest]

Drawing with DNA: 'Bioart' illuminates genomics

21 Jul 2005 Medical News Today

On any given day, tens of thousands of biologists around the globe

run DNA sequences of unknown function through a lightning-fast online

algorithm called BLAST - typically submitting 200 to 400 base pairs,

or " letters " of genetic code, to be matched against the billions of

letters for known genes. Searching for similarities that can shed

light on functional or evolutionary relationships, scientists

routinely use BLAST to churn through and produce vast amounts of

data. Everyday applications include genetic medicine and

pharmaceuticals. Yet this process and, more generally, genomics

remain dimly understood by the public.

" Ecce Homology, " an interactive " bioart " installation to be showcased

at SIGGRAPH 2005 - in Los Angeles, July 31 through Aug. 4 - quite

literally makes BLAST and genomics visible.

Headed up by new-media artist Ruth West - director of visual

analytics and interactive technologies at the University of

California, San Diego National Center for Microscopy and Imaging

Research and research associate with the UCSD Center for Research and

Computing in the Arts - the " Ecce Homology " project is an ongoing

collaboration among 11 biologists, artists and computer scientists

from UCSD, UCLA and the University of Southern California.

Named after Friedrich Nietzsche's Ecce Homo, a meditation on how one

becomes what one is, the project explores human evolution by

examining similarities - a.k.a. " homology " - between genes from human

beings and a target organism, in this case the rice plant.

" We are living in a time when we are generating enormous amounts of

genetic data, " said West, who trained as a microbiologist and began

her career in medical genetics. " But data is not knowledge - it's not

even information. A key concept of 'Ecce Homology' is to make an

important subject like genomics accessible to the general public. "

" Ecce Homology " uses a combination of dynamic media, computer vision

and computer graphics to visualize genomic data.

Custom software turns genes - incomprehensibly long strings of As,

Cs, Ts and Gs - into luminous pictograms that resemble Chinese or

Sanskrit calligraphy. Based on currently available biophysical

information, the pictograms are scientifically accurate

representations of proteins encoded for by the genes.

In the SIGGRAPH installation, the representations are rendered in a

40-foot wide and 12-foot tall space by five video projectors, with

the figures for human genes/proteins shown along a vertical axis and

for the rice along a horizontal.

A whole-body computer vision interface tracks the movements of

visitors and allows them to interact with the installation. By moving

their bodies slowly within the space, visitors can draw shimmering

light-filled traces. When a trace sufficiently matches a pictogram in

the human dataset, it triggers a real-time bioinformatics comparison:

BLAST begins to run, searching through the rice data for a homologue -

conducting in a novel (and visible) way the same sequence analysis

done by scientists. Results are presented as two superimposed

pictograms.

" This high-dimensional visualization reduces the complexity of

sequence codes to the sorts of shapes or patterns that a human being

can make sense of, " West said. " It is an artistic approach to

extracting what's important. And it is also an exploration of what

art might have to offer for discovery in the sciences. "

" Ecce Homology " premiered in 2003 at the UCLA Fowler Museum of

Cultural History.

At SIGGRAPH 2005, " Ecce Homology " is being showcased as part of the

international conference's Art Gallery and its Emerging Technologies

program. It will also be featured in the August 2005 issue of

Leonardo, an art, science and technology journal from MIT Press.

The project is supported by Intel Corp., NEC Solutions America/Visual

Systems Division and groups at UCSD, UCLA and USC.

To learn more: insilicov1.org

For more about SIGGRAPH 2005: siggraph.org/s2005

University of California - San Diego

http://www.ucsd.edu