Solving a 50-Year-Old Puzzle With AI

In 1972, the American biochemist, Christian Anfinsen, won the Nobel Prize in Chemistry for his work on ribonuclease. In his acceptance speech, he famously stated that, theoretically, a protein’s amino acid sequence should fully determine its structure. This hypothesis launched a 50-year quest to computationally predict a protein’s 3D structure based solely on its 1D amino acid sequence. Earlier this year, a new advancement in AI, called AlphaFold, finally solved this five-decade-long problem. Let’s take a look at what AlphaFold is, how it was trained, and what capabilities it offers. 

What is AlphaFold? 

AlphaFold is an AI system developed by Google’s DeepMind that is able to predict the structures of almost every protein made by the human body. This is very important because proteins are vital building blocks of living organisms and understanding their shapes could open the door for critical advancements in medicine. AlphaFold replaces traditional, more time-consuming techniques of working out protein structures which include X-rays, crystallography, cryogenic electron microscopy (Cryo-EM), and others. All of these methods required at least six months of work for researchers to come up with just one structure and thanks to AlphaFold it only takes a few minutes. 

How Was AlphaFold Trained? 

AlphaFold was trained on publicly available data consisting of ~170,000 protein structures from the protein data bank together with large databases containing protein sequences of unknown structures. However, the raw data still needed to be annotated to better train the AI system. The annotation methods required here include semantic segmentation, polylines, and others. Quality data annotation helps the AI system make more accurate predictions of the underlying physical structure of the protein and is able to determine highly accurate structures in a matter of days.

What are the Real-World Implications of AlphaFold? 

As we mentioned earlier, AlphaFold significantly shortens the time required for working out protein structures, but it also allows us to solve protein structures that researchers were working on for many years. AlphaFold also opens the door for new breakthroughs in medications. For example, researchers are working on a cure for Chagas disease, a life-threatening illness that can lead to heart failure and affects between six and seven million people. A team from the Geneva-based Drugs for Neglected Diseases initiative (DNDi) has found a molecule that appears to be capable of binding to a protein on Trypanosoma cruzi, the parasite that causes Chagas disease. Researchers can use DeepMind to study the protein’s structure to understand exactly how well a new drug is stopping the parasite from functioning. 

Another interesting application of AlphaFold is using it to find new biodegradable enzymes that can help us better manage pollutants, like plastic and oil, to break them down faster in ways that are more eco-friendly. PET, a strong plastic commonly used in bottles, takes hundreds of years to break down, but a new enzyme, called PETase, can start breaking down the material in a couple of days. In fact, work on this has already begun and AlphaFold can help expedite new advancements in this area. 

It is worth pointing out that AlphaFold was used in the battle against the ongoing COVID-19 pandemic. It was used by researchers to better understand some of the proteins in the novel Coronavirus. This information helped researchers better understand how the virus functions and ultimately find a cure. 

What’s Next for AlphaFold? 

While AlphaFold is already showing impressive results, there is still so much information that we need to learn from it. This includes things like how proteins form complexes, how they interact with DNA, RNA, or small molecules, and how we can determine the precise location of all amino acid side chains. In addition to this, researchers also need to understand how to use the discoveries made by AlphaFold in the development of new medicines, ways to manage the environment, and more.

Trust Mindy Support With All of Your Medical Data Annotation Needs

At Mindy Support, we understand the importance of highly accurate data annotation in the healthcare industry, which is why we have rigorous QA processes in place to ensure the highest levels of accuracy. We currently have 2,000 employees all over Ukraine and in other geographies globally to cover all required language skills. Our size and location allow us to source and recruit the needed number of candidates within a short time frame and we can scale your team quickly as well. Contact us today to learn more about how we can help you.