protein structure confidence metric · BEGINNER LESSON
pLDDT
pLDDT, or predicted local distance difference test, is a per-residue score that measures how confident AlphaFold2 is about the local structure of each amino acid in a protein. It is scaled from 0 to 100, where higher values usually mean the predicted local shape is more reliable.
What pLDDT measures
pLDDT is a local confidence score for proteins. “Per-residue” means it gives one value for each amino acid in the chain. It asks how well the predicted local structure would agree with an experimental structure, using a 0 to 100 scale. Higher numbers mean higher confidence and usually better local accuracy. [0,2]
[1] pLDDT: Understanding local confidence | AlphaFold - EMBL-EBI [3] Highly accurate protein structure prediction with AlphaFold
How it is related to lDDT-Cα
pLDDT is based on lDDT-Cα, which stands for local distance difference test using C-alpha atoms. C-alpha atoms are backbone atoms used as a simple reference point for protein shape. Unlike methods that first superimpose two structures, lDDT-Cα checks whether local distances are correct, so it can judge accuracy without alignment. [0,2]
[1] pLDDT: Understanding local confidence | AlphaFold - EMBL-EBI [3] Highly accurate protein structure prediction with AlphaFold
How to read the score
Scores above 90 are usually taken as very high local accuracy, where both backbone and side chains are typically predicted well. Scores above 70 usually suggest a correct backbone but some side-chain mistakes. Scores below 50 often appear in flexible or intrinsically disordered regions, or in places where AlphaFold2 lacks enough information. [0,1]
[1] pLDDT: Understanding local confidence | AlphaFold - EMBL-EBI [2] www.ebi.ac.uk · strengths-and-limitations-of-alphafoldStrengths and limitations of AlphaFold 2 | AlphaFold - EMBL-EBI
Why scores vary along one chain
A protein can contain regions that are easier or harder to predict. Globular domains often get higher pLDDT because they are more conserved and give the model more information to work with. Linkers between domains, disordered segments, or regions that change when binding a partner or a modification occurs often get lower scores. [0,1]
[1] pLDDT: Understanding local confidence | AlphaFold - EMBL-EBI [2] www.ebi.ac.uk · strengths-and-limitations-of-alphafoldStrengths and limitations of AlphaFold 2 | AlphaFold - EMBL-EBI
Limitations
pLDDT measures local confidence, not the relative position or orientation of different domains in the full protein. A high pLDDT across all parts does not mean AlphaFold2 knows how those parts are packed together. AlphaFold2 also has limits with point mutations, highly variable sequences, binding partners such as nucleic acids or small molecules, and membrane orientation. [0,1,2]
[1] pLDDT: Understanding local confidence | AlphaFold - EMBL-EBI [2] www.ebi.ac.uk · strengths-and-limitations-of-alphafoldStrengths and limitations of AlphaFold 2 | AlphaFold - EMBL-EBI [3] Highly accurate protein structure prediction with AlphaFold
SEE THE IDEA
How pLDDT works in practice
A chain of amino acids to be predicted
AlphaFold2 estimates the shape near each residue
A 0–100 confidence value for that residue
Higher means more trust in local shape, lower means caution
The sequence is predicted residue by residue, and pLDDT summarizes how confident each local prediction is.
pLDDT is a local confidence measure, not a whole-structure packing score.
TRY IT YOURSELF
Test your understanding
Choose an answer, then check the explanation.
PUT IT TOGETHER
Match each concept to its meaning
Drag a label onto its description, or select a label and then select a description. This also works with a keyboard.
4 pairs to match
Sources and further reading
This general lesson was generated from public source material. It explains a concept; it does not interpret a private paper or provide medical advice.