Seeing the shape of life.

Dates and assessment

Source published

2021-07-15

Bright published

2026-09-05

Substantive update

2026-09-25

Evidence state

Demonstrated

Independent verification

Not established by this source review

Last source review

2026-09-25

The claim in context

The human problem

A protein’s shape helps explain its function, but determining that shape can be difficult.

The prior constraint

Experimental structure determination required specialized instruments and substantial work.

AI’s actual role

A neural network integrated sequence and structural information to predict three-dimensional structure.

The documented result

The 2021 AlphaFold2 paper reported strong performance in the blind CASP14 assessment. On September 24, 2026, AlphaFold Database separately added high-confidence predictions covering 2,812 viral proteomes: 5,279 heterodimers and 2,749 homodimers, plus 4,681 high-confidence homodimers from the separate Viral Assembly Database collection.

Why it may matter

Researchers gained another way to investigate biological mechanisms. The later viral-pair resource gives pandemic-preparedness teams research hypotheses to inspect, prioritize and test experimentally.

Limitations

• Predictions carry uncertainty. A predicted structure does not establish function, interaction in a living system, vaccine efficacy or treatment benefit.
• The viral release is a research resource of predicted dimers, not experimentally validated structures; it does not model glycans, larger assemblies, the effects of genetic variation or host–pathogen interactions.
• The p53 view remains a static single-chain prediction, not an experimental structure or a simulation of protein folding.
• The p53 model's pLDDT values describe local confidence and do not establish the arrangement of distant regions.

Original evidence

• Highly accurate protein structure prediction with AlphaFold · Nature · paper
• AlphaFold DB human p53 AF-P04637-F1 model v6 · dataset
• AlphaFold Database adds viral protein complexes · EMBL · institution
• AlphaFold Database frequently asked questions · dataset
• AlphaFold Database viral protein complex collection · dataset

Attribution

Credit Bright AI Future and link the canonical Bright record.

• Link to the canonical Bright record.
• Keep material limitations with the claim they qualify.
• Link to the original evidence when repeating a substantive claim.
• Do not describe a source check or organization-reported result as independent verification.

Linked source material, quotations, trademarks and media remain subject to their owners’ terms. No reuse right is granted for third-party media.