---
id: protein-structure-prediction
title: "Protein Structure Prediction: AlphaFold, RoseTTAFold, and AI-Driven Structural Biology"
schema_type: article
category: ai
language: en
confidence: medium
last_verified: "2026-05-28"
created_date: "2026-05-24"
generation_method: ai_structured
ai_models:
  - claude-4.5-sonnet
derived_from_human_seed: true
conflict_of_interest: none_declared
is_live_document: false
data_period: static
completeness: 0.85
atomic_facts:
  - id: af-ai-protein-structure-prediction-1
    statement: AlphaFold demonstrated highly accurate protein structure prediction with deep learning.
    source_title: Highly accurate protein structure prediction with AlphaFold
    source_url: https://www.nature.com/articles/s41586-021-03819-2
    confidence: medium
  - id: af-ai-protein-structure-prediction-2
    statement: >-
      RoseTTAFold uses a three-track neural network for predicting protein structures and
      interactions.
    source_title: Accurate prediction of protein structures and interactions using a three-track neural network
    source_url: https://www.science.org/doi/10.1126/science.abj8754
    confidence: medium
  - id: af-ai-protein-structure-prediction-3
    statement: >-
      The CASP14 assessment evaluated the performance of protein structure prediction methods,
      including major deep-learning advances.
    source_title: Assessment of protein structure prediction in CASP14
    source_url: https://doi.org/10.1002/prot.26237
    confidence: medium
primary_sources:
  - id: ps-ai-protein-structure-prediction-1
    title: Highly accurate protein structure prediction with AlphaFold
    type: academic_paper
    year: 2021
    institution: Nature
    url: https://www.nature.com/articles/s41586-021-03819-2
  - id: ps-ai-protein-structure-prediction-2
    title: Accurate prediction of protein structures and interactions using a three-track neural network
    type: academic_paper
    year: 2021
    institution: Science
    url: https://www.science.org/doi/10.1126/science.abj8754
  - id: ps-ai-protein-structure-prediction-3
    title: Assessment of protein structure prediction in CASP14
    type: academic_paper
    year: 2021
    institution: Proteins
    url: https://doi.org/10.1002/prot.26237
known_gaps:
  - Predicting intrinsically disordered protein regions and dynamics
  - De novo design of enzymes with novel catalytic functions not found in nature
disputed_statements: []
secondary_sources: []
updated: "2026-05-28"
---
## TL;DR
Protein Structure Prediction: AlphaFold, RoseTTAFold, and AI-Driven Structural Biology: Protein structure prediction estimates three-dimensional protein structure from amino-acid sequence or related biological information.

## Core Explanation
AI structure prediction advanced sharply with deep learning systems evaluated in CASP. Strong public claims should distinguish structure prediction from experimental structure determination and from downstream functional validation.

## Further Reading

- [Highly accurate protein structure prediction with AlphaFold](https://www.nature.com/articles/s41586-021-03819-2)
- [Accurate prediction of protein structures and interactions using a three-track neural network](https://www.science.org/doi/10.1126/science.abj8754)
- [Assessment of protein structure prediction in CASP14](https://doi.org/10.1002/prot.26237)