posted 06 October 2011 13:05            

BillieGene.zip Human-Chromosome-Codons-SingleOutput.zip

As part of a tryout with data driven music, I thought I'd tackle sonifying the Human Genome - well part of it at least

The human genome consists of 23 chromosomes made from the genetic materials DNA or RNA. Two of which, the X and Y chromosomes are sex-determining (whereby XX expresses Female and XY Male). The human X chromosome is built out of more than 153 million base pairs (two nucleotides) of either an GC or AT pair. Each GC base pair consists of Guanine and Cytosine and the AT base pair is made up of Adenine and Thymine. This initial look at audio from the genome concentrated on the Y chromosome as the number of base pairs is 'only' 58 million base pairs!

The data is fed into a Quartz Composer composition which matches groups of nucleotides which form a three-nucleotide Codon. The data was pre-processed using the Unix commands : cat Homo_sapiens.GRCh37.64.dna.chromosome.Y.fa | sed 's/.../&,/g' > Y.txt in order to add a comma between each codon rather than having Quartz Composer compute this on each run. This had been previously tested as follows:

echo 'GCTCGGGACGCTGAGGTGGGAGGATTGCTTGAGCCCAGGAGTTTGACACCA
GCATGGGCAATATGGTAAGACCCCATTATCTACAAAAA' | sed 's/.../&,/g'

==>

GCT,CGG,GAC,GCT,GAG,GTG,GGA,GGA,TTG,CTT,GAG,CCC,AGG,AGT,TTG,ACA,CCA,
GCA,TGG,GCA,ATA,TGG,TAA,GAC,CCC,ATT,ATC,TAC,AAA,AA

Codons specify particular amino acids, which are used by living cells to encode proteins. According to the Codon in the sequence, the Codon is matched within the Quartz Composer composition to its amino acid which then maps this to a numeric value and output via OSCulator into Kyma. Several codons can encode the same amino acid, for example, the Codons TGT and TGC correspond to Cysteine whereas only TGG corresponds to the amino acid Tryptophan.

In Kyma these numeric values where then used to both select sound processing paths and also values for selected parameters. The output was then recorded into Ableton Live where I then carried out a kind of natural selection of the sounds I liked.

Attached are the Quartz Composer composition used to parse the data and output mapping data together with an example Kyma sound I used to create audio using the data. Also included is the OSCulator file, which takes the input value range and distributes them between 0 and 1

Lastly here's a movie of what came out (Drums were added btw) :
http://www.youtube.com/watch?v=kEEqJkYqmoI

Chromosome Data from : http://www.ensembl.org/Homo_sapiens/Info/Index

In the ftp link (from clicking on 'Download Human genome sequence' @ http://www.ensembl.org/Homo_sapiens/Info/Index [guest/guest works fine]) I used Homo_sapiens.GRCh37.64.dna.chromosome.X.fa.gz and Homo_sapiens.GRCh37.64.dna.chromosome.Y.fa.gz

[DNA/Chromosome infos from wikipedia]