The discovery of biological sequences encoded in DNA and RNA molecules has been a quantum leap for the life sciences. The sequences not only offer us a view on the genomic construction plans of all known life forms, they also allow researchers to monitor processes in the cell which is crucial for an understanding of the molecular dynamics in living organisms. The basic measurement tool of the modern biologist is the sequencer which transforms the macromolecular structures into some human and machine readable format which we know as long strings from a limited alphabet of four letters. It is likely that in the near future the data produced by sequencers will exceed in volume any other data generating technology. On the other hand a slightly different concept of a sequencer is well-known in electronic music. Here, the sequencer is an essential tool to generate repetitive structures that make up the technical aesthetics of electronic music since more than five decades. From the pioneers of classical electronic music to the variety of contemporary styles, musicians have always used the sequencer as a central instrument to produce rhythmical flow. Therefore the idea is to bring together the worlds of biology and music, using sequences as the basic structure that is naturally shared by both domains.
We have realized a "biomusical" instrument that is based on a
musical representation of biological sequence motifs. A
motif defines a sequence pattern that arises from a multiple
alignment of closely related sequences. The sequences are
particular realizations of the pattern and the motif
characterizes the variation of the evolutionary and/or
functionally related sequences. Motifs can be represented by a
positional weight matrix which defines a particular weight for
each possible letter at each pattern position. The weights
usually correspond to the observed frequencies of the letters and
provide a natural measure of importance that can be translated to
a musical weight of a particular note or beat. Merging the
concepts of step sequencers
the idea was to build a deterministic (non-random) generator of
musical phrases that uses given chords and four letter sequences
for the melodic component together with the positional weights
for the dynamics.
In our attempts to implement a performance oriented instrument we came across the x0xb0x project which combines an analog bass synthesizer with a programmable microcontroller (ATmega162) for sequencing. For this architecture we implemented a novel operating system (OS) which we refer to as "bioseq". The constraints for the bioseq OS were hard since there are only 16kb of program memory, not much in contrast to actual notebooks which can provide a million times more space for the code. Ending up with a few 1000 lines of efficient C code the final bioseq x0xb0x sequencer not only controls the internal bass synthesizer but also plays external sound generators like drum machines and synthesizers (see video below) via the integrated MIDI interface.
For a first test we utilized the ribosome binding site motif of E.coli K-12. In this experiment the bioseq program was fed with DNA sequences corresponding to length 16 fragments from the upstream region of experimentally verified gene starts.
All experimental results can be inspected here.
First experiments have indicated that biological sequences provide a rich reservoir of musically meaningful structures that can directly be used to drive electronic instruments like synthesizers and drum machines. In particular, the use of biological motifs in terms of a realtime controllable pattern generator creates a new musical instrument perfect for captivating live performances. Thereby, in terms of robustness, ease of use and fun factor the dedicated x0xb0x solution clearly outperforms a fragile and cumbersome general purpose computer-based realization. Future work will have to explore the huge variety of biological sequences from a broader range of functional roles and different organisms. The inclusion of high quality biological databases and sophisticated data mining tools will be an important step to systematically exploit the potential of biological motifs for generating rhythm-oriented electronic music.