Today I’ve been working on Magnificat and Nunc Dimits based around the base sequence of the chromosome of bateriophage phiX174 (the first complete mapped genome). On first inspection the code appears rather unpromising.
gagttttatc gcttccatga cgcagaagtt aacactttcg gatatttctg atgagtcgaa…
However the appearance is quite striking.
I’ve re-mapped the sequence to span an octave and imply rhythmic duration, and, oddly with a little bit of tweaking the system it comes out with some unexpectedly beautiful results. The table can be read like any ordinary table. Along the top is the genome ‘letter’ and down the side is how many times it comes in succession eg. aaaaaa would relate to A x 6.
| A | C | G | T | |
| 1 | D | A# | G | D# |
| 2 | C | B | E | F |
| 3 | A# | C | C# | G |
| 4 | G# | D | A | G# |
| 5 | F# | F | A | |
| 6 | E | B | ||
| 7 | C# |
According to the table this would result in the note E being sounded. Duration is also mapped by the number of repetitions although a lot more simply, the above Ax6 example would mean that note would last six quavers Tx3 for the quavers and so on. This gives us the original opening of the string:
gagttttatcgcttccatgacgcagaagttaa…
The string is the re-transcribed to show individual ‘events’:
g a g tttt a t c g c tt cc a t g a c g c a g aa g tt aa…
the letter and repetition data is then mapped and transformed as per the grid above giving:
G1 D1 G1 G#4 D1 D#1 A#1 G1 A#1 F2 B2 D1 D#1 G1 D1 A#1 G1 A#1 D1 G1 C2 G1 F2 C2…
This can be then translated into standard notation by simply taking the letter and # to indicate pitch and the number to signify the duration in units of a quaver.
The notation shows the resulting opening bars. The resulting monophonic line will form the basis of the organ part, tempo dynamics, rests and bar lengths can then be composed around this as the choral lines and text are worked into this sub-structure.
Currently this has been realised relatively simply. I have pasted the genome data into Word and converted it with a find and replace function to give the G1 D1 G1 G#4 D1 D#1 A#1 G1 A#1 F2 B2 D1 D#1 G1 D1 A#1 G1 A#1 D1 G1 C2 G1 F2 C2… string. I then put together a quick program/patch in Max MSP to convert the raw text of the word file into MIDI data and hence to audible pitches.
To give…
This audio is just the first few minutes of it, hopefully soon I’ll also work out ways of working in other dimensions to extract chords, dynamics and orchestrations out of it too. I have also been rolling out Markov chain stuff to provide a form of mutation to further work the data, but more of that anon.
Philosophically and ideologically I’m quite excited by the idea of combining the word setting and the DNA form. There feels as if there is something a bit subversive and yet also praising about the idea, a tug of war between evolutionary chemistry and the work of divine precision and creation. I can’t quite put my finger on it but working the piece out feels like it may be one amazing ride!
Ed Wright's Blog 