MIT researchers translated the coronavirus into music offering a new perspective for studying the virus

• Researchers at the Massachusetts Institute of Technology aren't only looking at the novel coronavirus, SARS-CoV-2, through a microscope, they are hearing it in musical form.
• Markus Buehler, material scientist and professor of engineering at MIT, described to Business Insider how he and his lab translated the structure of the coronavirus protein into a musical composition.
• Not only is it a new perspective into observing the virus, but Buehler said future applications of this data could help design antibodies that could counteract human infection with the coronavirus, which causes the respiratory illness COVID-19.

Amid an unprecedented pandemic, researchers have been rushing to understand the novel coronavirus, which causes the respiratory illness known as COVID-19.

While some are taking the traditional route of visually looking at the virus, dubbed SARS-CoV-2, under a microscope, researchers at the Massachusetts Institute of Technology are using another sense to study the coronavirus: hearing.

Markus Buehler, a material scientist and professor of engineering at MIT, told Business Insider that his lab, which specializes in the study of biological materials, had taken the approach of sonification — the process of conveying data in audio format — to protein materials, namely SARS-CoV-2. "When we go to the nano scale, we're actually looking at, basically, the atoms at the scale of individual molecules," Buehler said. "We realized that matter is always moving and vibrating because every atom is continuously shaking, so because of that, we can make any chemical structure into a sound."

His lab had previously applied sonification to materials like spiderweb silk and organ cells and soon became interested in creating a musical representation of the coronavirus protein.

Using machine learning and AI technology, Buehler said he and his team were able to "calculate the vibrational spectrum and make that audible," thus creating overlapping "melodies" that would ultimately create a composition or audible representation of the protein.

"When we translate the protein into sound and music, we basically follow the sequence by which these proteins are constructed — you can imagine this being a piano with 20 keys," Buehler said. "Each key has a unique sound, but this sound isn't a sound that comes from a string or a drum membrane or a trumpet."

---

"It's a sound that actually comes from the real vibrations off of chemical building blocks," he continued. Apart from being a new approach to observing the virus, Buehler told Business Insider that he believes there is a scientific application that could bring us closer to a solution to the pandemic.

"It has a real scientific application because once we have a description of matter and sound, ... these can be set into an artificial neural network," he said. "We can listen to it as humans, and begin to understand how proteins sound like and how mutations sound like."

Buehler added that it is easier to hear abnormalities in the structure of a protein rather than scrutinizing data and looking at it under a microscope at the nano-level.

"It's ... incredible to hear the kind of structures in there that you don't really see," he said. "If you just look at the letters, you don't see it. But if you listen to it, you can hear incredible detail in there."

Looking to the future, Buehler said the sonification research could potentially be used to design a protein that can counteract the coronavirus.

"In the case of coronavirus, we were interested in developing antibiotic candidates," Buehler said. "So we're feeding the information about how the spike protein is built, and we're trying to ask the artificial neural network to design a protein that fits like a lock-and-key into the spike protein, so the spike protein can no longer meet or match the human cell during the infection, but we'd rather bind into the antibody."

Hannah Davis, a programmer who works with data sonification as well and was not part of Beuhler's research, worked on an algorithm transposing literature into music. She told Business Insider that she thought Buehler and his lab were able to decide on pleasant notes to map different parts of data, but what is interesting to her is the functional purpose of their study.

"What's particularly interesting about this use case, to me, is less the music and more the functional sonification component — that they're using it to find a particular deviation in the pattern," she wrote in an email to Business Insider. "To do that they'll have to train their ears to learn the 'legend' of the sonification — to know what 'sounds normal' versus what might be worth paying better attention to."

"It's a way to almost subconsciously pick up on deviations from the data after listening to enough of it, she continued. "I'm not sure it's necessarily a path to a cure, but it's a very interesting reason to use sonification, and I'll be curious to see what comes of it!"

Read the original article on Business Insider