# google-deepmind/alphafold3 architecture

> Architecture diagram of the GitHub repository google-deepmind/alphafold3, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-29). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/google-deepmind/alphafold3
- Repository: https://github.com/google-deepmind/alphafold3

## Overview

AlphaFold 3 is a biomolecular-structure prediction inference package. Its principal workflow accepts molecular input, optionally searches sequence and template databases, prepares model features, runs neural-network inference, and post-processes predictions. The README documents data-pipeline and inference as independently switchable stages. Edges are limited to README statements and references visible in the SOURCE INDEX; unsampled implementation details and output-format wiring are not asserted.

Main entry points: [`src/alphafold3/data/pipeline.py`](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/data/pipeline.py) (MSA pipeline), [`src/alphafold3/model/features.py`](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/features.py) (Feature preparation), [`src/alphafold3/common/folding_input.py`](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/common/folding_input.py) (Fold input), [`src/alphafold3/data/templates.py`](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/data/templates.py) (Template search)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_input["Input and chemistry"]
  node_fold_input["Fold input<br/>[folding_input.py]"]
  node_chemistry["Component chemistry"]
end

subgraph group_search["Sequence and templates"]
  node_msa_pipeline["MSA pipeline<br/>[pipeline.py]"]
  node_msa_tools["Alignment tools"]
  node_template_search["Template search<br/>[templates.py]"]
  node_data_parsers["Alignment parsing<br/>[parsers.py]"]
end

subgraph group_inference["Feature and model inference"]
  node_feature_build["Feature preparation<br/>[features.py]"]
  node_atom_layout["Atom layout<br/>[atom_layout.py]"]
  node_model_pipeline["Inference pipeline<br/>[pipeline.py]"]
  node_model["Neural model<br/>[model.py]"]
  node_confidence["Confidence scoring<br/>[confidences.py]"]
end

subgraph group_structure["Structure processing"]
  node_post_process["Prediction post-processing<br/>[post_processing.py]"]
  node_structure_parse["Structure parsing<br/>[parsing.py]"]
  node_structure_model["Structure operations<br/>[structure.py]"]
  node_mmcif["mmCIF support<br/>[mmcif.py]"]
end

node_researcher(("Prediction user"))
node_sequence_db[("Sequence databases")]
node_template_db[("Structure database")]
node_weights["Model parameters<br/>[params.py]"]
node_result["Predicted structure"]

node_researcher -->|"supplies input"| node_fold_input
node_fold_input -->|"enters search"| node_msa_pipeline
node_msa_pipeline -.->|"runs alignments"| node_msa_tools
node_msa_pipeline -.->|"searches sequences"| node_sequence_db
node_msa_pipeline -->|"gets templates"| node_template_search
node_template_search -.->|"searches structures"| node_template_db
node_template_search -->|"uses parsers"| node_data_parsers
node_msa_pipeline -.->|"parses alignments"| node_data_parsers
node_fold_input -->|"supplies molecular input"| node_feature_build
node_feature_build -->|"uses atom layout"| node_atom_layout
node_feature_build -->|"uses components"| node_chemistry
node_feature_build -->|"uses MSA and templates"| node_msa_pipeline
node_feature_build -->|"provides features"| node_model_pipeline
node_model_pipeline -.->|"runs inference"| node_model
node_model -.->|"loads parameters"| node_weights
node_model -.->|"produces confidence data"| node_confidence
node_model_pipeline -.->|"post-processes prediction"| node_post_process
node_post_process -.->|"returns prediction"| node_result
node_structure_parse -->|"parses mmCIF"| node_mmcif
node_structure_parse -->|"builds structures"| node_structure_model
node_structure_model -->|"uses mmCIF utilities"| node_mmcif

click node_fold_input "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/common/folding_input.py"
click node_chemistry "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/constants/chemical_components.py"
click node_msa_pipeline "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/data/pipeline.py"
click node_msa_tools "https://github.com/google-deepmind/alphafold3/tree/main/src/alphafold3/data/tools"
click node_template_search "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/data/templates.py"
click node_data_parsers "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/data/parsers.py"
click node_feature_build "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/features.py"
click node_atom_layout "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/atom_layout/atom_layout.py"
click node_model_pipeline "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/pipeline/pipeline.py"
click node_model "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/model.py"
click node_weights "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/params.py"
click node_confidence "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/confidences.py"
click node_post_process "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/post_processing.py"
click node_structure_parse "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/structure/parsing.py"
click node_structure_model "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/structure/structure.py"
click node_mmcif "https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/structure/mmcif.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_fold_input,node_chemistry,node_researcher toneBlue
class node_msa_pipeline,node_msa_tools,node_template_search,node_data_parsers,node_sequence_db,node_template_db toneAmber
class node_feature_build,node_atom_layout,node_model_pipeline,node_model,node_confidence toneMint
class node_post_process,node_structure_parse,node_structure_model,node_mmcif toneRose
class node_weights,node_result toneIndigo
```

## Components

### Input and chemistry

- [Fold input](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/common/folding_input.py): `src/alphafold3/common/folding_input.py`
- [Component chemistry](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/constants/chemical_components.py): `src/alphafold3/constants/chemical_components.py`

### Sequence and templates

- [MSA pipeline](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/data/pipeline.py): `src/alphafold3/data/pipeline.py`
- [Alignment tools](https://github.com/google-deepmind/alphafold3/tree/main/src/alphafold3/data/tools): `src/alphafold3/data/tools`
- [Template search](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/data/templates.py): `src/alphafold3/data/templates.py`
- [Alignment parsing](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/data/parsers.py): `src/alphafold3/data/parsers.py`

### Feature and model inference

- [Feature preparation](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/features.py): `src/alphafold3/model/features.py`
- [Atom layout](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/atom_layout/atom_layout.py): `src/alphafold3/model/atom_layout/atom_layout.py`
- [Inference pipeline](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/pipeline/pipeline.py): `src/alphafold3/model/pipeline/pipeline.py`
- [Neural model](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/model.py): `src/alphafold3/model/model.py`
- [Confidence scoring](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/confidences.py): `src/alphafold3/model/confidences.py`

### Structure processing

- [Prediction post-processing](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/post_processing.py): `src/alphafold3/model/post_processing.py`
- [Structure parsing](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/structure/parsing.py): `src/alphafold3/structure/parsing.py`
- [Structure operations](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/structure/structure.py): `src/alphafold3/structure/structure.py`
- [mmCIF support](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/structure/mmcif.py): `src/alphafold3/structure/mmcif.py`

### Other components

- **Prediction user**
- **Sequence databases**
- **Structure database**
- [Model parameters](https://github.com/google-deepmind/alphafold3/blob/main/src/alphafold3/model/params.py): `src/alphafold3/model/params.py`
- **Predicted structure**

## Connections

- Prediction user → Fold input: supplies input (evidence: `README.md`)
- Fold input → MSA pipeline: enters search (evidence: `src/alphafold3/data/pipeline.py`)
- MSA pipeline → Alignment tools: runs alignments
- MSA pipeline → Sequence databases: searches sequences
- MSA pipeline → Template search: gets templates (evidence: `src/alphafold3/data/pipeline.py`)
- Template search → Structure database: searches structures
- Template search → Alignment parsing: uses parsers (evidence: `src/alphafold3/data/templates.py`)
- MSA pipeline → Alignment parsing: parses alignments
- Fold input → Feature preparation: supplies molecular input (evidence: `src/alphafold3/model/features.py`)
- Feature preparation → Atom layout: uses atom layout (evidence: `src/alphafold3/model/features.py`)
- Feature preparation → Component chemistry: uses components (evidence: `src/alphafold3/model/features.py`)
- Feature preparation → MSA pipeline: uses MSA and templates (evidence: `src/alphafold3/model/features.py`)
- Feature preparation → Inference pipeline: provides features (evidence: `src/alphafold3/model/pipeline/pipeline.py`)
- Inference pipeline → Neural model: runs inference
- Neural model → Model parameters: loads parameters
- Neural model → Confidence scoring: produces confidence data
- Inference pipeline → Prediction post-processing: post-processes prediction
- Prediction post-processing → Predicted structure: returns prediction
- Structure parsing → mmCIF support: parses mmCIF (evidence: `src/alphafold3/structure/parsing.py`)
- Structure parsing → Structure operations: builds structures (evidence: `src/alphafold3/structure/parsing.py`)
- Structure operations → mmCIF support: uses mmCIF utilities (evidence: `src/alphafold3/structure/structure.py`)
