What if you could set biology to music?
That’s a concept explored by Professor Markus Buehler, who has been doing a lot of work around what some might call modern alchemy, and has some pretty interesting theories on music.
In conjunction with the MIT/IBM Watson AI Lab, Buehler has looked at human proteins and genomic data, as he ponders the ways that music and matter align.
Why the focus on proteins?
“Proteins are the bricks and mortar that make up our cells, organs, and body,” Buehler explained to Kim Martineau for MIT News in 2020, when his coronavirus venture was brand new. “Alpha helix proteins are especially important. Their spring-like structure gives them elasticity and resilience, which is why skin, hair, feathers, hooves, and even cell membranes are so durable. … With IBM, we’re trying to harness this biochemical trait to create a protein coating that can slow the spoilage of quick-to-rot foods like strawberries.”
Another of his most prominent projects is his idea of putting the coronavirus to music.
“Our brains are great at processing sound! In one sweep, our ears pick up all of its hierarchical features: pitch, timbre, volume, melody, rhythm, and chords,” Buehler told Martineau, explaining why it’s interesting to combine something like a covid pathogen and a tune. “We would need a high-powered microscope to see the equivalent detail in an image, and we could never see it all at once. Sound is such an elegant way to access the information stored in a protein. Typically, sound is made from vibrating a material, like a guitar string, and music is made by arranging sounds in hierarchical patterns. With AI we can combine these concepts, and use molecular vibrations and neural networks to construct new musical forms. We’ve been working on methods to turn protein structures into audible representations, and translate these representations into new materials.”
Looking at the coronavirus’s spike protein structure, he suggested, his findings underscore the crafty nature of the virus itself, which wreaked havoc on the world’s population in 2020.
“It tricks our ears in the same way the virus tricks our cells,” he said. “It’s an invader disguised as a friendly visitor.”
And, he said, it sounds good, too: he called the result a “pleasant, even relaxing, tone,” but suggested that there’s real data encoded in the melody, too, and it’s also useful in research.
As for the methodology, Buehler explained that scientists trained a deep learning model on something called the ‘protein data bank’ that holds data on around 120,000 proteins, synthesized a protein in the lab, and built a process called end-to-end prediction in which a model shows a protein’s theoretical structure and assembly.
Setting these structures to music, Buehler argues, is helpful, because we can hear a lot more than we can see – in other words, that transfer means we are aligning the data with our most capable sense by translating it into sound.
“Translating proteins into sound gives scientists another tool to understand and design proteins,” Buehler said. “Even a small mutation can limit or enhance the pathogenic power of SARS-CoV-2. Through sonification, we can also compare the biochemical processes of its spike protein with previous coronaviruses, like SARS or MERS.”
You can also read more of his thoughts on applications and ramifications for processes like drug discovery.
All of this suggests we may want to ‘strike up the band’ when it comes to doing biological research and understanding what’s behind various kinds of molecular builds.
Look for more as I continue to report some of the most exciting things happening in the AI world today.
