Scroll Top

How AlphaFold is helping in treatment of protein misfolding diseases

Alphafold protein misfolding diseases: Just last month, AlphaFold launched their Protein Structure Database, a data set of high-quality predictions (95% accuracy) for the shape of every protein in humans, as well as proteins of 20 additional organisms that are vital for scientific research. This announcement has been lauded as the most important advancement since the mapping of the human genome.

Since a protein’s function depends on its shape, the key question at the AlphaFold project is “how do proteins fold up”? Historically, determining protein structures was a time-consuming and painstaking process. By utilizing AI, AlphaFold can accurately predict the shape of proteins, in mere minutes with atomic accuracy.

Putting this dataset into the hands of scientists will accelerate research that will advance our understanding of these essential building blocks of life, advance life-saving cures for diseases, and accelerate drug discovery.”

You can hear more about the AlphaFold story here:

What is AlphaFold technology?

AlphaFold is an artificial intelligence computer program developed by Google’s DeepMind. It was created in 2016 to try and solve the protein folding problem, a problem that refers to the inability to accurately determine the structure of a protein from an amino acid sequence. Using a deep learning system, AlphaFold is the first technology ever that has been able to predict the shape of a protein from an amino acid sequence down to atomic level accuracy.

How is AlphaFold technology helping in resolution of protein misfolding?

In July of 2021, AlphaFold made their platform open-source, effectively releasing the 3D structures of over 350,000 proteins. This doubles the understanding of the human proteome and allows for a better understanding of disease states and drug design. Gain Tx’s drug discovery platform SEE-Tx™,uses the 3D structure of a protein to find novel allosteric binding sites that when targeted with a small molecule will correct protein misfolding. This announcement has doubled the potential protein targets for Gain Tx and opens the door for additional therapies for protein misfolding diseases.

How accurate is AlphaFold protein structure prediction?

The AlphaFold program can predict the shape of a protein from a given amino acid sequence in just a few minutes, with atomic level accuracy. AlphaFold earned a median score of 92.4 out of 100, which is the level of accuracy comparable to industry standard techniques like X-ray crystallography.