A protein's sequence describes its chain of amino acids. Structure prediction estimates how that chain is arranged in three dimensions, providing a model that can complement experimental measurements.
For example, a researcher may start with a protein sequence and examine a predicted structure when planning an experiment. The prediction is a scientific starting point, not a complete explanation of the protein's behavior in every condition. Confidence can vary across regions of a predicted structure.
ELI5
Protein structure prediction tries to work out the shape of a protein from information about its building blocks.
For example, knowing the order of beads on a long string does not show exactly how the string sits in space. A computer model estimates that arrangement for a protein.
Is a predicted structure an experimental measurement?
No. It is a computational estimate that can support and complement laboratory investigations.
Are all parts of a prediction equally certain?
No. Different regions can have different confidence, so the output needs careful interpretation.

