DeepMind’s AlphaFold2 Solves 50-year Old Protein Fold Challenge

DeepMind has managed to solve decades-old problem with its machine learning model which opens up new avenue for AI-based predictions on protein sequencing and advanced research.

DeepMind’s ambitious AI-based AlphaFold system has managed to solve a complex cellular biology challenge that puzzled genome researchers for over 50 years. The DeepMind team published its first-ever AlphaFold Database. This Database provides open source access to protein structure predictions for 21 organisms, including humans. The new database would help researchers accelerate their work in the field of molecular biology and pharma research. Mapping of protein structures using 3D models built on DeepMind’s cutting-edge AI technology could provide massive trove of information related to human proteome and much more.

The 50-year Old Protein Fold Challenge

Proteins are the building blocks of life. There are two main principles driving the nature of protein once can synthesize in a lab—amino acids and the 3d structure. While we know more or less how amino acids work in chains, the 3D structuring remained a big puzzle for years. Researchers were unable to solve a unique challenge in how proteins fold in shapes. This problem is famously called as “Protein Folding Problem’ or PFP. Scientists at DeepMind solved the PFP puzzle by advancing their AI model, driving it through AlphaFold project.

What is DeepMind’s AlphaFold Project?

AlphaFold is DeepMind’s AI system used as a forecasting tool to map the 3d structure of proteins based on amino acid sequencing. DeepMind co-partnered with EMBL’s European Bioinformatics Institute (EMBL-EBI)to develop the AlphaFold Database (DB). The information available in AlphaFold can be freely accessed and modified for the benefit of scientific community engaged in the protein sequencing and genetic mapping research.

Organizers of Critical Assessment of protein Structure Prediction (CASP) recognized DeepMind AI system AlphaFold as a solution to the grand PFP challenge.

AlphaFold2 demonstrates the strong correlation between computational work in the field of biology and advanced molecular research. This could further assist in improving accuracy of predictions used to understand the nature and structure of important protein classes, including membrane proteins. By bringing together biology, physics and machine learning, DeepMind intends to positively influence the penetration of AI and computing for advanced sciences like drug design and environmental sustainability.