JMM 8, Winter 2009, section 5
Fredrik Søegaard & Claus Gahrn
SOUNDMAPPING THE GENES
“Art is the sedimentation of human misery.” (Adorno 2004).
The above remark by Adorno implies that art has its own way of preserving human experience, using a metaphor from geology and zoology, as well as its own languages for so doing (painting, musicmaking, sculpturing, etc).
Biology also has its own way of preserving some of the information about living beings and a language for so doing – the DNA coding sequences present in all biological beings.
SMTG is a project involving music composition and improvisation based on biological data – the complete DNA genetic code of the H1 Histonine protein of the rainbow trout.
Nature has always been one of the dominant aesthetic ideals for artists of any art form. Painters have used nature in their work from cave paintings up until our day, and composers and musicians have dedicated numerous works to the celebration of nature (see Adorno 2004).
In music, however, there is an ongoing discussion regarding exactly how nature is relevant in a musical context. What does a sunset sound like? Or a sea?
In 1986 Japanese-American Biologist Susumu Ohno from the Beckman Research Institute of The City of Hope, Duarte, California, created a system for composing music from DNA code sequences, transcribing the four nucleotides into the diatonic scale according to a set of rules formulated by Ohno (Ohno 1986). SMTG is a contemporary project in this tradition, using the DNA coding sequence from one protein of the rainbow trout to form e.g. the melodic structure, the rhythmic, the dynamic, etc. Structures in the 642-nucleotide-long H1 Histonine protein code sequence are transcribed into the chromatic scale and transferred into the MIDI (Musical Instruments Digital Interface) language. Thus the code sequence can be used as a chromatic melody or as MIDI information, controlling chosen musical parameters in electronic and digital environments. In this way, nature itself can appear as a ‘controller’ – via pitch or MIDI information – of the music in the form of the structural characteristics of the genetic code sequence. In addition to providing pitch information and creating melody and/or MIDI information to control electronic parameters, this complete H1 Histonine coding sequence also brings a specific form to the music, as this coding sequence is different from every other existing sequence.
5.2. About the Coding Sequence Translation
The translation from the genetic code to music is of course one of the basic issues of interest within SMTG. Ohno (Ohno 1986) suggests a translation based on the diatonic scale. The problem is that this scale consists of seven notes, but the DNA code only has four different ‘notes’. Ohno solves the problem by including the octave and thus arrives at eight notes: two notes per nucleotide. This entails an element of choice as the composer at any point in the coding sequence has to choose one of two notes. Furthermore, adding to the choices, there are several diatonic scales: Ionian,Dorian, Phrygian etc. (as well as melodic and harmonic minor and other synthetic modes), and they are all asymmetrical – half and whole steps are distributed unevenly, an aspect which has no immediate correlate in the four nucleotides.
Four note scales can be found in music, though: the tetrachords of Ancient Greek music theory, for instance. These also exist in several forms, however, according to how the half/whole steps are distributed, as all of the classic tetrachords – CDEF, DEFG, EFGA, FGAH, GAHC, AHCD, HCDE – apart from the Lydian one from the note F, consist of one half step and two whole steps. Since they are all assymetrical as well, the mapping process would consequently involve an element of choice regarding which tetrachord to use.
For SMTG we wanted a system of translation that was unambiguous and completely symmetrical and it had to be based on the number of nucleotides. We found that if one included the next nucleotide to a given one (ie.CG) you would arrive at twelve possibilities:For each of the four nucleotides it can be followed by one of three others apart from itself, that is 4 x 3 (= 12). If it is followed by an identical nucleotide, the translated note just doubles its length and this also adds rhythmic variety to the coding sequence melodies. This makes a completely unambiguous translation into the chromatic scale possible. And, furthermore, the chromatic scale is one out of two music scales that are fully symmetrical (the other one being the whole tone scale).
This translation is used for generating melodies in a simple fashion from various coding sequences, but it is also transferred into MIDI language, thus enabling code sequence structures to be used as controllers of chosen parameters in electronic music equipment (see below).
5.3. About DNA ‘Rhythm’
SMTG also exploits the concept of rhythm based on mechanisms in the DNA. The genetic code, generating information on the basis of which amino acids are to form a specific protein, is based on a triplet reading of the RNA strings, and triplets are also occurring in musical rhythms. The protein synthesis can thus be seen (or rather heard) as a waltz, a mazurka or any other piece of music organized in units of three beats each (see Søegaard 2003).
Research into the linguistics of nucleotide sequences (Brendel et al. 1986) studies the concept of ‘words’ in continuous languages – languages devoid of blanks – and introduces an operational definition of words. By means of this strategy, nucleotide sequences may become the objects of linguistic analysis. The typical word size of the nucleotide language is found to range from 3 to 7 (tri- to heptamers). As different genomes have distinct vocabularies, comparisons of these vocabularies can serve as a basis for revealing functional and evolutionary relatedness of sequences.
For each protein code sequence, it is possible to decide which polymers are ‘words’ and which are to be avoided – as mentioned above, different genomes have distinct vocabularies. Linguistic analysis will clarify the ‘word’ polymers and this reading will result in ‘sentences’ consisting of different polymers, from trimers to heptamers. This framing of different ‘word’ polymers results in an asymmetrical rhythm, resembling rhythm concepts from e.g. North Indian classical music, also used in Western classical music by composers like Olivier Messiaen.
As the result of this linguistic analysis, the nucleotide sequences may be read as a series of asymmetrically rhythmic measures.
In this section we describe four pieces of music, all of which have been performed in concert. Video documentation is presented for the first two.
H1 HISTONINE RAINBOW TROUT CODING SEQUENCE
Melodic Improvisation (Guitar, MIDI genemap, percussion, electronics)
The solo melody that appears from the beginning is the H1 protein code in its melodic form using the chromatic translation of the code mentioned above.
The following guitar-harmonies use the MIDI translation of the code to control various dynamic filterings of the sound. These filterings are then also used to control the percussion instruments towards the end of the composition.
H1 HISTONINE RAINBOW TROUT CODING SEQUENCE
Rhythmic improvisation (guitar, MIDI genemap, percussion, electronics)
This piece starts off with the H1 protein melody played in a very high tempo. The guitar, with pitch controlled by the MIDI genemap, and percussion improvise in various sections throughout the composition.
At the end, the super-fast version of the H1 melody is heard again. The breaks in the improvised section use rhythmic framing principles corresponding to nucleotide sequence linguistics taken from various DNA sources.
HUMAN X PRIMORDIAL HEPTAMER POLYRHYTHMS (Percussion/Electronics)
This composition is made up of percussion improvisation upon an electronically recorded matrix of multi-layered digital percussion instruments. The layers form a complex polyrhythmical structure consisting of nucleotide sequence linguistic readings of genetic code-sequences, the polymers originating from above mentioned linguistic analysis, taken from excerpts of a variety of DNA strings. The percussionist is then instructed to use similar techniques employing polyrhythms and improvise in the same “language” as the strings.
20 PROTEINS ACROSS 5 TIME POINTS(Electronics)
This piece is based upon protein data , and not, as in the previous pieces, upon DNA codes. The data was given to us by Professor Mustapha Kassem of Odense University Hospital who wished to be able to gain new insights into the protein data by listening to it as sound or music. We decided to make three versions using different approaches to mapping the protein data into parameters that would be easier for the ear to comprehend. The parameters in this case were the changes in pitch, tempo and loudness of pitch. We did not wish the music to reflect a particular style or genre, although the result in the end could be said to come close to certain abstract classical works from the 20th century.
With growing access to biological algorithms, it is becoming possible to let nature be a part of music-making in the form of data-interfaces between electronic sound, musical instruments and complex computer software based on biological information, such as the MIDI genemap used in SMTG. This opens the possibility for a closer relationship between biological information, structures and forms in music, allowing for completely new ways of conceiving questions of musical material and form – instead of sonata form you could have protein-based form and instead of three-part fugues you could have polymer readings of DNA strings, just to name a few examples.
The research into the linguistics of nucleotide sequences suggests an interesting relationship with the growing field of work in biosemiotics Together these areas will surely result in added knowledge about the “words” and “sentences” in the nucleotide sequences. By comparing one system of semiotics (spoken and written languages) with other systems (of, say, music and biology) one might even produce a better knowledge of how all of the systems work.
Future goals of the project will then be to make more music, using biological information, and to integrate the project with scientific research in the various affiliated areas: molecular biology, sonification of DNA information and biosemiotics.
SMTG thus becomes a truly cross-disciplinary project, involving art (music), the natural sciences (biology), nucleotide sequence linguistics (linguistics) and technology (sonification) – not a bad accomplishment.