AlphaFold for Protein Folding
AlphaFold (also known as AlphaFold2, AF2, or AF) is an advanced AI model developed by DeepMind for predicting the 3D structures of proteins from the amino acid sequence. It is a transformer-based method with excellent, state-of-the-art performance, winning the CASP14 competition.
How to use AlphaFold for Protein Structure Prediction
Inputs
- Protein Sequence
A string representing the single-letter amino acid sequence, consisting of the common 20 amino acids. This input is intended for single-chain (monomeric) proteins only.
Outputs
- PDB File
File containing the predicted 3D coordinates of each residue in the input sequence.
- pLDDT Scores
Specified in the b-factor column of the PDB file. A per residue confidence score between 0 and 100 with higher being better.
- PAE Scores
A JSON file containing the residue-pair confidence scores. Smaller is better.
Examples
Due to the speed of AlphaFold, Copilot serves an alternative, called ESMFold.
Fold ‘DIHICGICKQQFNNLDAFVAHKQSGSQ’
