Musical Proteins

Protein sequences turned into audible melodies
28 June 2019




Protein sequences can be translated into music which allows scientists to understand them better, and create new types of proteins.

Proteins are the building blocks of all living things, making up everything from the human eye to spider webs to jelly. They come in all shapes and sizes, from tightly wound corkscrews, to loose “noodle-like” structures, each with its unique function.

Every protein is made up of some combination of a family of chemicals called amino acids. There are just 20 unique amino acids, but by rearranging them in different ways, like beads on a string, we can make all the various proteins.

The first protein structures were mapped out in the 1960s. Yet, we still don’t understand exactly which structures lead to different functions.

Now, Markus Buehler and his colleagues at MIT, have found a way to turn each amino acid into a unique chord on a scale. This allows them to translate a complicated molecule into a sound: something simple, which is easy to rearrange and play with.

“We have translated this language of proteins into a different space, into the space of audible information,” says Buehler. “By listening to the proteins, we can begin to link the structure of proteins, how they fold, what kind of function they have, with how they sound.”

This means they can start to understand how the structure of the protein relates to its function.
and gain a better insight into diseases which are caused by protein mutations. They can also start to design new proteins with specific functions.

Beyond potential medical uses, this also opens up a new avenue for material scientists. Proteins are strong, flexible materials which are also renewable and biodegradable. Developing them as a source of new materials is a promising step towards a more sustainable future.

What’s the science behind this? Do amino acids really make sounds?

At the nanoscale, everything is in constant motion. This means that if we zoom into an amino acid molecule, it will be vibrating at a certain set of frequencies.

Buehler and his team found that each type of amino acid had a unique vibration signature. They transposed this to an audible frequency to come up with the 20 tone scale.

They then trained an artificial neural network (AI)  to learn these notes, and the combinations of notes which make up known proteins.

Using this, the AI can start making slight variations to the melodies, and by running their converting program backwards,these melodies can then be used as blueprints to create new types of proteins with different functions.

For now, these are still simply blueprints. The team have been modelling them using computer simulations but they are yet to synthesise them in the lab. It’s early days, but Buehler and the team “hope to possibly find proteins that are even better than the ones nature has designed for us.”

Want to hear how an amino acid sounds? Search on Google Play for Amino Acid Synthesiser to download an app which allows you to play with the sounds yourself.


Add a comment