Scientists Use AI to Create Music through Proteins

It turns out that proteins have unique musical vibrations.
Fabienne Lang

It's an exciting day in the worlds of physics and music. Researchers have discovered that proteins have their own unique musical vibrations, and have created their very own musical scores by using AI and proteins. 

Proteins are the building blocks of the human body, found in our hair, our tendons, our muscles, and everywhere else. They're a part of us, and it turns out they can make music too. 

The findings were published in the journal APL Bioengineering on Tuesday.


When nature and science work together, they can create music

Proteins are naturally found in anything from silk to human cells, however, scientists have no way to automatically figure out their design, and need them to create new proteins. 

Scientists Use AI to Create Music through Proteins
Using musical scores to code the structure and folding of proteins composed of amino acids, each of which vibrates with a unique sound, Source: Markus J. Buehler/MIT

This is where engineers, Markus Buehler from MIT, and Chi Hua Yu from the National Cheng Kung University in Taiwan, come in. The researchers discovered that each of the twenty amino acids that make up a protein has its very own sound. 

Moreover, the team put two and two together and noticed that, like music, proteins have a hierarchical structure. For instance, the basic structuring of a protein is due to the varied ordering of its amino acids, which is very similar to how music scores are structured. 

So Buehler and Yu decided to take their experiment further by assigning each protein trait a musical analogue. For example, a specific amino acid was assigned the musical note "C", and so on and so forth. The result? A musical stave of protein notes

Buehler then created a whole raft of musical protein pieces that he's shared on SoundCloud.

AI and musical proteins

But that wasn't enough to satiate Buehler and Yu's thirst for science and technology. The team then trained a deep learning algorithm on data sets that included a number of proteins and their linked musical scores. 

Once the training was done, the algorithm was prompted with bits of music to start producing entirely new musical compositions. This is music like you've never heard before. 

It's not yet certain how this discovery will assist science, however, Buehler is optimistic as he stated "This paves the way for making entirely new biomaterials. Or perhaps you find an enzyme in nature and want to improve how it catalyzes or come up with new variations of proteins altogether."

Add Interesting Engineering to your Google News feed.
Add Interesting Engineering to your Google News feed.
message circleSHOW COMMENT (1)chevron
Job Board