Tragedy Transposed, The Sounds of HIV

November 12, 2010 8:06 AM

There is no question that HIV is an ugly virus in terms of human health. Each year, it infects some 2.7 million additional people and leads to some two million deaths from AIDS. But a new album manages to locate some sonic beauty deep in its genome. Sounds of HIV (Azica Records) by composer Alexandra Pajak explores the patterns of the virus's nucleotides as well as the amino acids transcribed by HIV, playing through these biologic signatures in 17 tracks.

Preview samples are available to listen to on "Classics Online", and another sample is located on the SciAm blog.


Some of the proceeds from the new album, which was released October 26 and is performed by the Sequence Ensemble, will go toward HIV vaccine research at the Emory Vaccine Center.


Genetics- and science-inspired music are genres unto themselves, and everything from proteins to meteorite compositions have also been translated—if not always melodiously—at least tonally. Pajak took as her basic formula the National Institutes of Health's record of the retrovirus' genome and the thousands of coded letters which get transcribed by an enzyme into DNA in a cell once it's infected. But her latest work goes a step farther than simply plunking out the nucleotide-based notes (A, C, G and D, which fills in for thymine) on the keyboard.

In addition to the basic base pairs, she explains, "I assigned pitches for the amino acids," which are manufactured once HIV enters a human cell. She ordered the 20 directly encoded amino acids based on their affinity for water (with arginine as the most hydrophilic and isoleucine the most hydrophobic) and gave them notes on the A minor scale based on this property. In the piece, Pajak occasionally layers amino acid and nucleotide phrases with different instruments to create musical interest, although she has kept the sequence of the notes true to the genome.

Pajak decided to compose the piece in a minor scale to acknowledge the profound sadness the virus causes. Her past DNA-based pieces include a profile of the West Nile virus. She became curious about the HIV genome, especially when its complete structure was sequenced to single-nucleotide resolution in 2009, and decided to explore it "just for myself to see what it would sound like."

posted by infinite intimation (20 comments total)

Also;

Ever since the 1970s, when Iannis Xenakis began composing music based on mathematics, architecture, and physics, science and music have been trying to find common ground. In October, University of Georgia graduate student Alexandra Pajak will become part of that rapprochement of music and science when she releases an album of original music that she composed based on the DNA of HIV, she told the Daily Scan. "Sounds of HIV is a musical translation of the genetic code of HIV," Pajak's liner notes read. "Every segment of the virus is assigned music pitches that correspond to the segment's scientific properties." Pajak has assigned certain notes and pitches to specific amino acids and nucleotides. The composition's Prelude and Postlude correspond to the first and last 100 nucleotides, and the sections named after the proteins (Proteins 1-9) represent translations of the amino acid sequences, she says. Pajak got the idea to compose DNA-based music when, as an undergrad, a genetics professor asked her to compose a symphony based on the DNA of the college founder's mother. At various conferences ever since, several researchers have commissioned CDs from her, she says. Pajak is currently recording with the Atlanta Symphony Orchestra, and part of the sales proceeds will go to the Emory Vaccine Center, which conducts HIV research.


Kinexus Bioinformatics has converted the amino acid sequences and corresponding DNA sequences of the following 12 cell signalling proteins (EGF Receptor, PI 3-Kinase, PTP-1B Phosphatase, VHI-related Phosphatase, PP2B/Calcineurin, Phospholipase A2, PI-specific PLC, cGMP Phosphodiesterase, Adenylate Cyclase, Gi-α1 G Protein, H-Ras, and Caspase 8) into musical notes to produce a unique melody for each protein. The soundtracks are based on the hydrophobicity scores of the amino acids in the primary structure of each protein; the more hydrophilic the amino acid, the higher the musical note that was assigned. The melody notes are of amino acids and background music is of the corresponding DNA. Different scales and tempos were used for the proteins and a selection of different instruments were also employed. The soundtracks were produced by Dr. William Campbell and are freely available for download from the following url: http://www.kinexus.ca/scienceTechnology/gallery/music/music.html




Kinexus has converted the amino acid sequences and corresponding DNA sequences of the following 12 cell signaling proteins (EGF Receptor, PI 3-Kinase, PTP-1B Phosphatase, VHI-related Phosphatase, PP2B/Calcineurin, Phospholipase A2, PI-specific PLC, cGMP Phosphodiesterase, Adenylate Cyclase, Gi-α1 G Protein, H-Ras, and Caspase 8) into musical notes to produce a unique melody for each protein. The soundtracks are based on the hydrophobicity scores of the amino acids in the primary structure of each protein; the more hydrophilic the amino acid, the higher the musical note that was assigned. The melody notes are of amino acids and background music is of the corresponding DNA. Different scales and tempos were used for the proteins and a selection of different instruments were also employed. The soundtracks were produced by Dr. William Campbell.

posted by infinite intimation at 8:08 AM on November 12, 2010

Previously, the amino-acid sequence of H1N1 Hemagglutinin as ambient music.
posted by James Scott-Brown at 8:12 AM on November 12, 2010 [1 favorite]

Humans at their best. Facing the indifferent cruelty of nature and transforming it into music.

Great post.
posted by Joe Beese at 8:16 AM on November 12, 2010

Anything that can help us understand this virus and make any aspect of it useful and beneficent is a plus in my book.
posted by blucevalo at 9:02 AM on November 12, 2010

I hope that, in the future, this music is played as the funeral march when the last extant copies of the then near extinct virus, now rendered impotent by a variety of vaccines and actual cures, are ceremonially gathered into a vial and thrown into a volcano.
posted by adipocere at 9:15 AM on November 12, 2010 [3 favorites]

Am I alone in finding this all beside the point? This whole business just underscores a belief held by many that scientists are more interested in silly puzzles than actual solutions. Now, I know this not to be true (and am fact a biologist myself) but seriously. Maybe HIV the musical can wait until after we've come up with a cure?
posted by eggyolk at 10:40 AM on November 12, 2010

Maybe HIV the musical can wait until after we've come up with a cure?

When there is a financial motivation for a cure, Big Pharma will find it.

Not until then.
posted by blucevalo at 11:11 AM on November 12, 2010

eggyolk: "Maybe HIV the musical can wait until after we've come up with a cure?"

This is a composer. The alternate was not her finding a cure, but making some piece of music based on the topology of the Himalayas or whatever. Or maybe you think EVERYONE should be finding a cure? What have you done today to find a cure for AIDS?
posted by idiopath at 11:51 AM on November 12, 2010 [2 favorites]

There is much leeway in the various approaches to data sonification (the audible analog to visualization). Whether the result comes out pretty or ugly or ambient or rock or whatever usually has much more to do with the sonification algorithm than it does with the data.

Xenakis would trust the feel of a musical piece entirely to the data - a true experiment. But this gets you music that audiences don't like, and given that a-priori the piece was limited to A minor she wasn't as trusting of the data as Xenakis would be.
posted by idiopath at 12:05 PM on November 12, 2010

Whether the result comes out pretty or ugly or ambient or rock or whatever usually has much more to do with the sonification algorithm than it does with the data.

Yes! Finally! Someone comes out and says it!

I would be interested in sonification using an extremely naive algorithm applied to a natural phenomenon that I thought would turn out musical using that algorithm (e.g. you could turn measurements of the height of a buoy into an audio signal really easily, but it would mainly have two components (waves and tides)). That would be pretty cool. I haven't done this yet because I haven't found the right phenomenon to reinterpret.
posted by a snickering nuthatch at 12:55 PM on November 12, 2010

"Maybe HIV the musical can wait until after we've come up with a cure?"

Eggyolk: Too late.

The "Butthole Duet" alone is...memorable. Though not as catchy as "Pop-a-Boner."


Seriously though, I'm looking forward to listening to the links in this post - when I'm at a computer that has speakers. *sigh*
posted by dendritejungle at 2:22 PM on November 12, 2010

There is much leeway in the various approaches to data sonification (the audible analog to visualization).

True: what claims to be a transcription turns out to be a translation 'inspired' by the data. Putting the data into a key certainly makes it more listenable, as does using conventional instruments, and we all have to make a living somehow. Plus, it's a worthy kind of marketing since the proceeds go towards funding research.

But this kind of thing usually upsets me as an audio-/music geek, because it obscures the very subject. I don't want to hear a natural phenomenon or structure shoehorned into the polite conventions of chamber music, and would rather just donate some money than be forced to sit there and listen to it. Like Jpfed, I want to hear a mechanical rendition of a natural thing from which I might be able to infer some new information.

Happily, not only were these protein sequences converted into musical form at the link i_i supplied above, all 12 are also offered in MIDI format, along with this key (pdf) enable the data to be easily reverse engineered, w00t.
posted by anigbrowl at 2:35 PM on November 12, 2010

I agree, also, books that deal with people getting over the murders of their loved ones should not be written until murder itself is eliminated, same for any kind of movie that deals with people dying from heart attacks, and there should be no more paintings that are dealing with conflict and strife in warzones. Sorry, that is not intended as deep sarcasm, and I don't mean to be mean, I appreciate your position, I guess I just deeply disagree with the sentiment (possibly because I am mistaken). I would agree that exploiting any one individual's life story would possibly be wrong, but this is a generalized thing, not like making a story about one person who dies, or suffers, but rather saying, people do die and suffer- but also live, and love beautifully.

'Cures', if they actually are even possible, will come out of some silly solution to a silly puzzle. We really shouldn't pander to people who could not care less about science. It is also possible that 'cures' will cause further unforeseen consequences. The reality is that there is no predicting when, where or how a cure will arrive... but until then, people living a life must goes on, must find ways to go on, and hopefully find people along the way that will care about their situation, or even think that someone they love can face the same situation.

Pajak, who is a graduate student in clinical social work at the University of Georgia and composes the music in her spare time.

But really, what of the idea of providing compassion, comfort, knowledge that people are thinking about this deadly killer... for a long time HIV/AIDs sat silently as a disease of stigma. The victims relegated to the farthest reaches of so much of societies collective mind, kept as far away as possible... in the fringes, on the outside, always 'the other', considering that this could bring the killer into the room, into people's homes, down to earth; manageable?

There is more to "science and healing" than simply "cures". Care is also key. Art can be Therapy, Treatment, Care. Therapists studying PTSD in Warfighters does not mean that other people are not also working on ways to end war.

Evelyn Glennie speaks to these concerns, teaching to listen with one's whole body. Music Therapy, musical solidarity. The cynical may still say "solidarity bah what is that good for!", well, solidarity is huge, and a shift from how it has been for decades.


This set of art though? She readily discusses the “massaged reality” of these notes… using real instruments, tempo, timing, phrasing, timbre, the things of which music is born. From my reading of the presentation of this music, there is no pretense that it is "the genes", or "scientific data", it is advertised and presented as Music. The author is not a biologists or a geneticist. Music is in the expression; until it is interpreted, it is, arguably, just notes.

In my experience, more people rally against robotic, cold-steel-like reality-musicality, I know I don't like a MIDI nearly as much as a phsyical instrument in almost all cases, I remember the response to an “ai musician” posted on the blue a year or more back, it was called cold, robotic, clinical, not worth the while (though I liked it)… I dunno, I am torn, I get the desire to have this UGLY UGLY killer not immortalized with beautiful notes and tempos and rhythms, and LIFE… and rather to be a flat, static, clinical-cold piece devoid of any (traditionally) beautiful hints or tones or themes… but, I like this as it is also. I am glad that the musicians infused the simple “notes” with life, and human expression. To me, it speaks of a future where people are still there to do that expression. I think to deny that there is beauty in the life of HIV victims, is to deny the humanity inherent in human life. Which must include suffering.

I am reminded of the chapter in Sartre's 'the wall' titled Erostratus (where HIV is Herostratus);

A story about a misanthropic man who resolves to follow the path of *Herostratus and make history by means of an evil deed -- in this case, by killing six random people (one for each bullet in his revolver). The man is exhilarated by the sense of power he receives when carrying his revolver on the streets within his pocket. "But I no longer drew assurance from that [the revolver], it was from myself: I was a being like a revolver, a torpedo or a bomb." Sartre gives the reader an insightful account about how a man's nature changes with the objects of his possession, but the object itself is unable to change the internal man, as seen in the conclusion.

*On July 21, 356 BC, Herostratus in his quest for fame set fire to the Temple at Ephesus in what is now Turkey. The temple was constructed of marble and considered one of the most beautiful buildings in the world. It had been built by King Croesus of Lydia to replace an older site destroyed during a flood and honoring a local goddess conflated by the Greeks with Artemis, their goddess of the hunt, the wild and childbirth. Measuring 130 metres long (425 feet) and supported by columns 18 metres high (60 feet), it was one of the Seven Wonders of the Ancient World.
Aftermath

Far from attempting to evade responsibility for his act of arson, Herostratus proudly claimed credit in an attempt to immortalize his name in history. To dissuade similar-minded fame-seekers, the Ephesean authorities not only executed him, but also condemned him to a legacy of obscurity by forbidding mention of his name under penalty of death. This did not stop Herostratus from achieving his goal, however, as the ancient historian Theopompus recorded the event and its perpetrator in his Hellenics.

All that is not to say that I, however, am not also glad that the original unmassaged data exists... just making a case for interpretive art in the face of science. Reality is never so simple that it can be expressed without interpretation. -it is just a matter of where on the continuum of interpretation something exists.
posted by infinite intimation at 11:28 AM on November 14, 2010

It wasn't the instrumentalists alone who made the data pretty. The prettiness was also built into the sonification algorithm (and the composer had a free hand with the harmonies, totally uninfluenced by the data). I am not talking about coldly taking the data with no transformation - some transformation is absolutely necessary for the data to have some sonic relevance in the resulting output.

But the existing piece sounds a bit too much as if Picasso had painted Guernica by tracing parts of a battlefield photo that made up a smiley face. I have a hard time believing, given her methods, that a totally random input would have sounded any different whatsoever. A proposal for future composition that claims to sonify some soft of data: there should be at least one movement or theme that is clearly identified for the audience as being random input to the same algorithm that later handles the actual data. Show me the difference between the data you are sonifying and noise, or you have not shown me anything whatsoever.
posted by idiopath at 3:50 PM on November 14, 2010

infinite intimation: "get the desire to have this UGLY UGLY killer not immortalized with beautiful notes and tempos and rhythms, and LIFE… and rather to be a flat, static, clinical-cold piece devoid of any (traditionally) beautiful hints or tones or themes… "

This has nothing to do with anything anyone said in this thread. Beautiful is relative, but if you claim to portray something, at least give me some small way to distinguish what it is you are portraying, what makes this data different from some other data or white noise or a markov sequence or whatever.
posted by idiopath at 3:59 PM on November 14, 2010

Isn't it that it is a translation of the letter semiotic sequence, the 'sequence' that represents a representation of HIV? That is all that makes this 'different' from pure random.

It was instrumentalists breathing life, but also, perhaps more so, the composer, in creating timing, phrasing, and layering, perhaps is where the uniqueness is arrived at. I don't get how it could "portray HIV" more accurately, it is transposition, or maybe "transliteration" is a better term for it; not portrayal of HIV.

I think what confuses me is that I feel like a 'totally random' input would sound different? No? I mean, I hear the 'random' patterns of the notes, but they represent letters, which represent chemical compounds, which represent a structure. From my understanding of what I read, it wasn't just an algorithm, but composition choices by Pajak that determined the "tempo and phrasing" of each note (like how long it was held)

"I just have to think up some instrumentation, some rhythms, the tempo," she adds

(it is possible she meant "...by using an algorithm", but it would seem that she has experience and skill at composition, and might not have left these choices up to an algorithm generator)

Postlude V. seems to me to display an interesting interplay between sparse and textured with interesting 'random' (not "Random") notes.

Genetic code has no 'letter' or symbol for "tone length", so I guess that is where the human hand comes in, not in the hands of the instrumentalists (for Pajak, but not for Xenakis). The arbitrary seems to come from that temporal element.

I guess I just meant me, personally, I can understand a desire to show the terrifying killer in a cold, clinical, harsh light.
Sorry, I think I came across as trying to prove something, or 'defend' this work (worth defending, I like it), but I guess I really didn't get across clearly that this is conflicting to me as well, the struggle to balance aesthetics with reality, and tonality with reality, how long notes are held seems to be under the control of the composer, perhaps arbitrarily?

On rereading and relistening to Xenakis' files, I think I see what you are saying (the notes are ALL the same length in those tracks [and I agree, it does seem to show that he has a confidence in the data to come out in a complex and interesting texture without adding anything other than fixed length notes {and I should note, I don't think his are simple/cold/clinical/ugly/at all} it has an interesting 'fluidity' to it] Xenakis' pieces actually are starting to closely remind me of some parts by Philip Glass in Koyaanisqatsi)

Is what you are saying about how Pajak has added in "note length" as an arbitrary feature?
It is entirely possible I just don't understand the sonification process well enough. Or I could also be misinterpreting your words. Apologies if information is being lost in transmission (if it is, I am pretty sure it is being misunderstood on my side of the equation).

I find both the techniques and technology as well as creativity involved fascinating, and beautiful (both the arbitrary tempo and rhythm, and layered parts of Pajak's creations, and the more sonically accurate pieces by Xenakis) I would like to learn more.
posted by infinite intimation at 7:49 PM on November 14, 2010

infinite intimation: "Genetic code has no 'letter' or symbol for "tone length""

The genetic code also has no 'letter' or symbol for degree within the scale A minor. The fact that the base pairs use letters just like music is a trivial coincidence, not a property of enzymes or sounds (though it may be a good starting point if you are analyzing various pedagogical traditions and their evolution of mnemonics and terminology).

Sonfication is much like visualization - just as one graph may choose to show a number as a height of a rectangle another may show it as the size of a pie slice or another may choose to show it as a color, one sonfication of a DNA sequence could use a base pair as a duration, another use it as a pitch, another as a chord, another as an amplitude. The mappings can be arbitrary but the goal in choosing is to in some way convey an otherwise harder to see property of the data - in that sense it is helpful to provide a key or use a method that you can presume the audience is familiar with. Unless the sonification element is not a meaningful property of the piece, and it is being used as a novelty or a parlor trick.

My point regarding randomness is that if an audience member doesn't know how to listen for where the information is, in other words where things turn out differently with a different input data, then one cannot know whether the data was a meaningful part of the result, or the algorithm is such that any data at all would sound the same to the human ear and the source of data was irrelevant.

infinite intimation: "Xenakis' pieces actually are starting to closely remind me of some parts by Philip Glass in Koyaanisqatsi"

Getting a bit OT: that is kind of like saying "Shakespeare's plays are starting to closely remind me of some parts of Checkov's Uncle Vanya". There may be a kernel of truth there but the ordering and comparison both seem a bit silly.
posted by idiopath at 9:26 PM on November 14, 2010

I do find it interesting that she transcribed the various enzymes ordered by affinity with water to individual notes - if for example she had kept this element and then tied durations and dynamics to base pairs and maybe used stop codons as key changes or some other continuity breaking shift, I would have a bit more respect for how this was done. Simply deriving a note from a base pair of the same name and filling so much in intuitively just seems simultaneously lacking in imagination and rigor. Every work must have its clinamen, but with the combination of her chosen algorithm and the room she gave herself for interpretation the clinamen crowds out the constraints to a pretty massive degree.
posted by idiopath at 9:58 PM on November 14, 2010

Sorry, I dropped a bit of jargon there that I should have defined in context: "Based on a conception of the movement of atoms in Lucretius' On the Nature of Things, the clinamen is the primordial anti-constraint: it makes creation possible by introducing chance and spontaneity in an ordered universe [Motte 1986b]."
posted by idiopath at 10:05 PM on November 14, 2010 [1 favorite]

A really interesting article discussing some of the basics of the "Digital Humanities/Humanities Computing" fields; unsworth's "what is Humanities Computing"

I'll give the short answer to the question »what is humanities computing?« up front: it is foreshadowed by my two epigraphs. Humanities computing is a practice of representation, a form of modeling or, as Wallace Stevens has it, mimicry. It is also (as Davis and his co-authors put it) a way of reasoning and a set of ontological commitments, and its representational practice is shaped by the need for efficient computation on the one hand, and for human communication on the other. We'll come back to these ideas, but before we do, let's stop for a moment to consider why one would ask a question such as »what is humanities computing?«

First, I think the question arises because it is important to distinguish a tool from the various uses that can be made of it, if for no other reason than to evaluate the effectiveness of the tool for different purposes. A hammer is very good nail-driver, not such a good screw-driver, a fairly effective weapon, and a lousy musical instrument. Because the computer is – much more than the hammer – a general-purpose machine (in fact, a general-purpose modeling machine) it tends to blur distinctions among the different activities it enables. Are we word-processing or doing email? Are we doing research or shopping? Are we entertaining ourselves or working? It's all data: isn't it all just data processing? Sure it is, and no it isn't. The goals, rhetoric, consequences, benefits, of the various things we do with computers are not the same, in spite of the hegemony of Windows and the Web. All our activities may all look the same, and they may all take place in the same interface, the same ›discourse universe‹ of icons, menus, and behaviors, but they're not all equally valuable, they don't all work on the same assumptions – they're not, in fact, interchangeable. To put a more narrowly academic focus on all this, I would hazard a guess that everyone reading this uses a word-processor and email as basic tools of the profession, and I expect that many readers are also in the humanities. Even so, you do not all do humanities computing – nor should you, for heaven's sake – any more than you should all be medievalists, or modernists, or linguists.
....
I. Humanities computing as model or mimicry

Davis et al. use the term »surrogate« instead of »mimicry« or »model«. Here's what they say about surrogates:

The first question about any surrogate is its intended identity: what is it a surrogate for? There must be some form of correspondence specified between the surrogate and its intended referent in the world; the correspondence is the semantics for the representation. The second question is fidelity: how close is the surrogate to the real thing? What attributes of the original does it capture and make explicit, and which does it omit? Perfect fidelity is in general impossible, both in practice and in principle. It is impossible in principle because any thing other than the thing itself is necessarily different from the thing itself (in location if nothing else). Put the other way around, the only completely accurate representation of an object is the object itself. All other representations are inaccurate; they inevitably contain simplifying assumptions and possibly artifacts.[4]

posted by infinite intimation at 11:07 PM on November 14, 2010 [1 favorite]