Understanding The AI Alphafold and It’s Significance In the Protein Folding Problem

Suchir Jey
4 min readDec 24, 2023

--

Since 1960, scientists have been researching ways to solve the protein folding problem. Recently, AI programs have made it possible to tackle the entire problem in every shape and form and to expand our current understanding and knowledge on the different protein structures. But how does artificial intelligence play a huge role in our proteins?

The Solution to Folding Proteins

All humans contain proteins, in fact every cell in the human body contains some sort of protein. On top of all this, we need to intake proteins for healthy living. Alphafold is a company that uses AI to predict protein structure and to solve protein folding. The AI program used by Alphafold is run by Deepmind.

Diagram of one structure of an unfolded and folded protein

What is the problem Alphafold is solving?

AlphaFold has pretty much solved how to fold single proteins in relation to the 50-year Protein Folding Problem. They managed to come up with a solution by using their AI program as previously mentioned. Protein folding happens because each protein has a string of beads called polypeptides made of different chemicals which are referred to as amino acids. The attraction and repulsion between the 20 different types of amino acids cause the string to curl up as a result of protein folding. Protein folding can cause the immune system to refrain production of antibodies for certain protein structures and protein folding can also lead to a variety of diseases like cancer. This becomes a problem because antibodies are what help fight off sicknesses and diseases from entering your body and without it, the human body can get sick and easily vulnerable to getting sick.

Diagram of Polypeptide chain

What is the history of Alphafold?

Alphafold is an AI program and a company that was created in a partnership with EMBL-EBIN. The first public release and announcement was on July 22, 2021, with the Alphafold 2 software but the Alphafold project began in 2016. On their release, they were already covering over 350,000 protein structures. On July 28, 2022, they covered over 1 million to well over 200 million structures today according to Google Deepmind. Alphafold has won the 2023 Albert Lasker Basic Medical Research Award for solving the challenge of predicting protein structures.

Company and AI program that created Alphafold

How does the technology work and how does it help scientists?

The AI platform was made to predict protein structures so we can learn, modify, and control proteins. The technology uses amino acids which are the beginning of the formation of proteins which are compared to known protein structures to find sequences. Then with those structures, multiple sequence alignments are created to provide an output of a prediction of a 3D protein structure. As mentioned earlier, protein folding can prevent antibodies from forming in that protein cell in your body which is bad since antibodies help kill bacteria and illnesses in our body and allow us to not get sick frequently as well as reducing the risk of diseases from protein folding. Therefore this will allow scientists to control proteins to prevent them from folding and modify them if they are misforming in order to reduce the risk of the consequences mentioned previously above. It will also give scientists the ability to gain further knowledge about particular proteins and how they work. This is how this technology can create an impact on scientists.

Photo From Alphafold of protein evolution using the Alphafold Database

Potential Future of Alphafold

Alphafold is slowly but shortly predicting and getting information on every human protein structure in existence. This technology has already had a massive impact and will continually create an impact in other areas in which proteins are involved. The next step for Alphafold is quite promising as they seem to be adventuring into other habitats and species protein models, especially ocean animals as Demis Hassabis one co-founder of Deepmind, talked about taking an exhibition to the Indian Ocean.

Key Takeaways

After reading this article, you now have an idea about how Alphafold uses Deepmind to make a difference in the world and can impact a large population in the world. Here are some main takeaways.

  • Alphafold predicts the way protein structures will look like before they are formed to reduce risk of folding
  • With this technology we can create a future where proteins aren’t a problem at all.

Thanks for reading my article. Please leave any feedback below! ☺️

--

--