FoldMatch

Version 0.4.2

Overview

FoldMatch is a Python toolkit to encode macromolecular 3D structures into fixed-length vector embeddings for efficient large-scale structure similarity search and clustering.

Reference: Multi-scale structural similarity embedding search across entire proteomes.

A web-based implementation using this tool for structure similarity search is available at rcsb-embedding-search.

If you are interested in training a new model with a new structure dataset, visit the rcsb-embedding-search repository, which provides scripts and documentation for training.

Features

Residue-level embeddings computed using the ESM3 protein language model
Sequence-based embeddings from FASTA files without requiring 3D structures
Structure-level embeddings aggregated via a transformer-based aggregator network
Fast and efficient FAISS-based similarity search
Structural clustering using the Leiden algorithm for biological assembly identification
Command-line interface implemented with Typer for high-throughput inference workflows
Python API for interactive embedding computation and integration into analysis pipelines
High-performance inference leveraging PyTorch Lightning, with multi-node and multi-GPU support

Installation

From PyPI

pip install foldmatch

From Source (Development)

git clone https://github.com/rcsb/foldmatch.git
cd foldmatch
pip install -e .

Requirements:

Python ≥ 3.12
ESM 3.2.3
Lightning 2.6.1
Typer 0.24.1
Biotite 1.6.0
FAISS 1.13.2
igraph 1.0.0
leidenalg 0.11.0
PyTorch with CUDA support (recommended for GPU acceleration)

Optional Dependencies:

faiss-gpu for GPU-accelerated similarity search (instead of faiss-cpu)

Usage

The package provides two main interfaces:

Command-line Interface (CLI) for batch processing and high-throughput workflows
Python API for interactive use and integration into custom pipelines

Command-Line Interface (CLI)

The CLI provides three main command groups: fm-structure for computing embeddings from a folder of structure files, fm-sequence for computing embeddings from protein sequences in FASTA files, and fm-search for building, updating, and querying FAISS databases for similarity search.

Structure Embedding Commands

`fm-structure residue`

Calculate residue-level embeddings using ESM3 from a folder of structure files. All chains in each structure are processed. Outputs are stored as PyTorch tensor files (default) or CSV files.

fm-structure residue \
  --src-folder data/structures \
  --output-path results/residue_embeddings \
  --structure-format mmcif \
  --batch-size 8 \
  --devices auto

Key Options:

--src-folder: Folder containing structure files (.cif, .pdb, or .bcif, including .gz variants)
--output-path: Directory to store embedding files
--output-format: separated (individual files) or grouped (single JSON)
--output-name: Filename when using grouped format (default: inference)
--write-tensor / --no-write-tensor: Write embeddings as torch tensor (.pt) files instead of CSV files when using separated format (default: disabled)
--structure-format: mmcif, binarycif, or pdb
--structure-file-extension: Override the default file extension used to filter structure files in --src-folder. Pass an empty string to disable extension filtering. When unset, the defaults for the chosen --structure-format are used.
--min-res-n: Minimum residue count for chain filtering (default: 0)
--batch-size: Batch size for processing (default: 1)
--num-workers: Data loader workers (default: 0)
--num-nodes: Number of nodes for distributed inference (default: 1)
--accelerator: Device type - auto, cpu, tpu, hpu, gpu (default: auto)
--devices: Device indices (can specify multiple with --devices 0 --devices 1) or auto
--strategy: Lightning distribution strategy (default: auto)

`fm-structure chain`

Compute chain-level embeddings from a folder of structure files. By default, residue embeddings are computed as a first step and stored in --res-embedding-folder, then aggregated into chain embeddings. Use --no-compute-residue-embedding to skip the residue step and use pre-computed residue embeddings.

# End-to-end: compute residue + chain embeddings
fm-structure chain \
  --src-folder data/structures \
  --res-embedding-folder results/residue_embeddings \
  --output-path results/chain_embeddings \
  --batch-size 4

# Using pre-computed residue embeddings (stored as .pt files)
fm-structure chain \
  --src-folder data/structures \
  --res-embedding-folder results/residue_embeddings \
  --output-path results/chain_embeddings \
  --no-compute-residue-embedding \
  --batch-size 4

# Using pre-computed residue embeddings stored as .csv files
fm-structure chain \
  --src-folder data/structures \
  --res-embedding-folder results/residue_embeddings \
  --output-path results/chain_embeddings \
  --no-compute-residue-embedding \
  --res-embedding-format csv \
  --batch-size 4

Key Options:

--src-folder: Folder containing structure files
--res-embedding-folder: Directory for residue embedding files (output when computing, input for chain aggregation)
--output-path: Directory to store chain embedding CSV files
--compute-residue-embedding / --no-compute-residue-embedding: Compute residue embeddings first (default: enabled)
--res-embedding-format: Format of the pre-computed residue embedding files when --no-compute-residue-embedding is set. Options: pt (torch tensor files) or csv (default: pt). Ignored when residue embeddings are computed on-the-fly.
--output-format: separated (individual files) or grouped (single JSON)
--output-name: Filename when using grouped format (default: inference)
--structure-file-extension: Override the default file extension used to filter structure files in --src-folder. Pass an empty string to disable extension filtering.
All other options similar to fm-structure residue

`fm-structure assembly`

Compute assembly-level embeddings from a folder of structure files. By default, residue embeddings are computed as a first step and stored in --res-embedding-folder, then aggregated into assembly embeddings. Use --no-compute-residue-embedding to skip the residue step and use pre-computed residue embeddings.

# End-to-end: compute residue + assembly embeddings
fm-structure assembly \
  --src-folder data/structures \
  --res-embedding-folder results/residue_embeddings \
  --output-path results/assembly_embeddings \
  --min-res-n 10 \
  --max-res-n 10000

# Using pre-computed residue embeddings (stored as .pt files)
fm-structure assembly \
  --src-folder data/structures \
  --res-embedding-folder results/residue_embeddings \
  --output-path results/assembly_embeddings \
  --no-compute-residue-embedding \
  --min-res-n 10 \
  --max-res-n 10000

# Using pre-computed residue embeddings stored as .csv files
fm-structure assembly \
  --src-folder data/structures \
  --res-embedding-folder results/residue_embeddings \
  --output-path results/assembly_embeddings \
  --no-compute-residue-embedding \
  --res-embedding-format csv \
  --min-res-n 10 \
  --max-res-n 10000

Key Options:

--src-folder: Folder containing structure files
--res-embedding-folder: Directory for residue embedding files (output when computing, input for assembly aggregation)
--output-path: Directory to store assembly embedding CSV files
--compute-residue-embedding / --no-compute-residue-embedding: Compute residue embeddings first (default: enabled)
--res-embedding-format: Format of the pre-computed residue embedding files when --no-compute-residue-embedding is set. Options: pt (torch tensor files) or csv (default: pt). Ignored when residue embeddings are computed on-the-fly.
--output-format: separated (individual files) or grouped (single JSON)
--output-name: Filename when using grouped format (default: inference)
--structure-file-extension: Override the default file extension used to filter structure files in --src-folder. Pass an empty string to disable extension filtering.
--min-res-n: Minimum residues per chain (default: 0)
--max-res-n: Maximum total residues for assembly (default: unlimited)
All other options similar to fm-structure residue

`fm-structure download-models`

Download ESM3 and aggregator models from Hugging Face.

fm-structure download-models

Sequence Embeddings Commands

`fm-sequence residue`

Calculate residue-level ESM embeddings from protein sequences in a FASTA file. No 3D structure information is required. Outputs are stored as PyTorch tensor files (default) or CSV files.

fm-sequence residue \
  --fasta-file sequences.fasta \
  --output-path results/residue_embeddings \
  --batch-size 8 \
  --devices auto

Key Options:

--fasta-file: FASTA file containing protein sequences
--output-path: Directory to store embedding files
--output-format: separated (individual files) or grouped (single JSON)
--output-name: Filename when using grouped format (default: inference)
--write-tensor / --no-write-tensor: Write embeddings as torch tensor (.pt) files instead of CSV files when using separated format (default: disabled)
--min-res-n: Minimum residue count for sequence filtering (default: 0)
--batch-size: Batch size for processing (default: 1)
--num-workers: Data loader workers (default: 0)
--num-nodes: Number of nodes for distributed inference (default: 1)
--accelerator: Device type - auto, cpu, tpu, hpu, gpu (default: auto)
--devices: Device indices (can specify multiple with --devices 0 --devices 1) or auto
--strategy: Lightning distribution strategy (default: auto)

`fm-sequence chain`

Compute chain-level embeddings from protein sequences in a FASTA file. By default, residue embeddings are computed as a first step and stored in --res-embedding-folder, then aggregated into chain embeddings using the transformer-based aggregator. Use --no-compute-residue-embedding to skip the residue step and use pre-computed residue embeddings.

# End-to-end: compute residue + chain embeddings
fm-sequence chain \
  --fasta-file sequences.fasta \
  --res-embedding-folder results/residue_embeddings \
  --output-path results/chain_embeddings \
  --batch-size 4

# Using pre-computed residue embeddings
fm-sequence chain \
  --fasta-file sequences.fasta \
  --res-embedding-folder results/residue_embeddings \
  --output-path results/chain_embeddings \
  --no-compute-residue-embedding \
  --batch-size 4

Key Options:

--fasta-file: FASTA file containing protein sequences
--res-embedding-folder: Directory for residue embedding tensor files (output when computing, input for chain aggregation)
--output-path: Directory to store chain embedding CSV files
--compute-residue-embedding / --no-compute-residue-embedding: Compute residue embeddings first (default: enabled)
--res-embedding-format: Format of the pre-computed residue embedding files when --no-compute-residue-embedding is set. Options: pt (torch tensor files) or csv (default: pt). Ignored when residue embeddings are computed on-the-fly.
--output-format: separated (individual files) or grouped (single JSON)
--output-name: Filename when using grouped format (default: inference)
All other options similar to fm-sequence residue

`fm-sequence download-models`

Download ESM3 and aggregator models from Hugging Face.

fm-sequence download-models

Search Commands

`fm-search build structures`

Build a FAISS database from structure files for similarity search. Residue embeddings are computed first using ESM3, then aggregated into chain or assembly embeddings.

fm-search build structures \
  --structure-folder data/pdb_files \
  --output-db databases/my_structures \
  --res-embedding-folder tmp \
  --granularity chain \
  --min-res 10 \
  --use-gpu-index

Key Options:

--structure-folder: Directory containing structure files
--output-db: Database path (prefix for .index and .metadata files)
--res-embedding-folder: Directory for intermediate residue embeddings
--structure-format: mmcif, binarycif, or pdb
--granularity: chain or assembly level embeddings
--file-extension: Filter files by extension (e.g., .cif, .bcif, .pdb)
--min-res: Minimum residue count (default: 10)
--use-gpu-index: Use GPU for FAISS index construction
--accelerator, --devices, --strategy: Inference device settings
--batch-size-res, --num-workers-res, --num-nodes-res: Residue embedding settings
--batch-size-aggregator, --num-workers-aggregator, --num-nodes-aggregator: Aggregator settings

`fm-search update structures`

Update an existing FAISS database with new or replacement structure files. Structures with IDs already present in the database are replaced; new IDs are added. The FAISS index is fully rebuilt after merging.

fm-search update structures \
  --structure-folder data/new_structures \
  --output-db databases/my_structures \
  --res-embedding-folder tmp \
  --structure-format mmcif \
  --granularity chain \
  --min-res 10 \
  --batch-size-res 8

Key Options:

--structure-folder: Directory containing new or updated structure files
--output-db: Path to the existing FAISS database to update
--res-embedding-folder: Directory for intermediate residue embeddings
--structure-format: mmcif, binarycif, or pdb
--granularity: chain or assembly level embeddings
--file-extension: Filter files by extension (e.g., .cif, .bcif, .pdb)
--min-res: Minimum residue count (default: 10)
--use-gpu-index: Use GPU for FAISS index construction
--accelerator, --devices, --strategy: Inference device settings
--batch-size-res, --num-workers-res, --num-nodes-res: Residue embedding settings
--batch-size-aggregator, --num-workers-aggregator, --num-nodes-aggregator: Aggregator settings
--log-level: Logging level - info, warn, or debug (default: info)

`fm-search build embeddings`

Build a FAISS database from a directory of pre-computed embedding files (.csv or .pt). The filename without extension is used as the embedding ID in the database. This is useful when embeddings have been previously computed with any of the fm-structure or fm-sequence commands.

fm-search build embeddings \
  --embedding-folder results/chain_embeddings \
  --output-db databases/my_structures \
  --file-extension .pt

Key Options:

--embedding-folder: Directory containing pre-computed embedding files (.csv or .pt)
--output-db: Database path (prefix for .index and .metadata files)
--file-extension: Filter by extension (.csv or .pt). If not specified, collects both
--use-gpu-index: Use GPU for FAISS index construction
--log-level: Logging level (default: info)

`fm-search update embeddings`

Update an existing FAISS database with new or replacement embeddings from pre-computed files (.csv or .pt). Embeddings with IDs already present in the database are replaced; new IDs are added.

fm-search update embeddings \
  --embedding-folder results/new_embeddings \
  --output-db databases/my_structures \
  --file-extension .pt

Key Options:

--embedding-folder: Directory containing pre-computed embedding files (.csv or .pt)
--output-db: Path to the existing FAISS database to update
--file-extension: Filter by extension (.csv or .pt). If not specified, collects both
--use-gpu-index: Use GPU for FAISS index construction
--log-level: Logging level (default: info)

`fm-search build sequences`

Build a FAISS database from protein sequences in a FASTA file. Residue embeddings are computed first using ESM3, then aggregated into chain embeddings. The FASTA sequence names are used as embedding IDs.

fm-search build sequences \
  --fasta-file sequences.fasta \
  --output-db databases/my_sequences \
  --res-embedding-folder tmp \
  --batch-size-res 4

Key Options:

--fasta-file: FASTA file containing protein sequences
--output-db: Database path (prefix for .index and .metadata files)
--res-embedding-folder: Directory for intermediate residue embeddings
--min-res-n: Minimum residue count for sequence filtering (default: 0)
--use-gpu-index: Use GPU for FAISS index construction
--accelerator, --devices, --strategy: Inference device settings
--batch-size-res, --num-workers-res, --num-nodes-res: Residue embedding inference settings
--batch-size-aggregator, --num-workers-aggregator, --num-nodes-aggregator: Chain embedding inference settings
--log-level: Logging level (default: info)

`fm-search update sequences`

Update an existing FAISS database with new or replacement embeddings computed from protein sequences in a FASTA file. Embeddings with IDs already present in the database are replaced; new IDs are added.

fm-search update sequences \
  --fasta-file new_sequences.fasta \
  --output-db databases/my_sequences \
  --res-embedding-folder tmp \
  --batch-size-res 4

Key Options:

--fasta-file: FASTA file containing protein sequences
--output-db: Path to the existing FAISS database to update
--res-embedding-folder: Directory for intermediate residue embeddings
--min-res-n: Minimum residue count for sequence filtering (default: 0)
--use-gpu-index: Use GPU for FAISS index construction
--accelerator, --devices, --strategy: Inference device settings
--batch-size-res, --num-workers-res, --num-nodes-res: Residue embedding inference settings
--batch-size-aggregator, --num-workers-aggregator, --num-nodes-aggregator: Chain embedding inference settings
--log-level: Logging level (default: info)

`fm-search query structure`

Search the database for structures similar to a query structure.

fm-search query structure \
  --db-path databases/my_structures \
  --query-structure query.cif \
  --structure-format mmcif \
  --granularity chain \
  --top-k 100 \
  --threshold 0.8 \
  --output-csv results.csv

Key Options:

--db-path: Path to FAISS database
--query-structure: Query structure file
--structure-format: mmcif or pdb
--granularity: chain or assembly search mode
--chain-id: Specific chain to search (optional)
--assembly-id: Specific assembly ID (optional)
--top-k: Number of results per query (default: 100)
--threshold: Minimum similarity score (default: 0.8)
--output-csv: Export results to CSV (optional)
--min-res: Minimum residue filter (default: 10)
--max-res: Maximum residue filter (optional)
--accelerator: Device type - auto, cpu, tpu, hpu, gpu (default: auto)
--use-gpu-index: Use GPU for FAISS search

`fm-search query embedding`

Search the database using a single pre-computed embedding file (.csv or .pt). The filename stem is used as the query ID. No model inference is required — the embedding is loaded directly and queried against the FAISS index.

fm-search query embedding \
  --db-path databases/my_structures \
  --embedding-file results/chain_embeddings/1acb.A.pt \
  --top-k 100 \
  --threshold 0.8 \
  --output-csv results.csv

Key Options:

--db-path: Path to FAISS database
--embedding-file: Pre-computed embedding file (.csv or .pt). The filename stem is used as the query ID
--top-k: Number of results to return (default: 100)
--threshold: Minimum similarity score (default: 0.8)
--output-csv: Export results to CSV (optional)
--use-gpu-index: Use GPU for FAISS search
--log-level: Logging level (default: info)

`fm-search query sequences`

Search the database using protein sequences from a FASTA file. Each sequence is used as a separate query, producing its own ranked result list. Residue and chain embeddings are computed first using ESM3, then each sequence is searched against the database.

fm-search query sequences \
  --db-path databases/my_structures \
  --fasta-file queries.fasta \
  --res-embedding-folder tmp \
  --top-k 100 \
  --threshold 0.8 \
  --output-csv results.csv

Key Options:

--db-path: Path to FAISS database
--fasta-file: FASTA file with protein sequences (each sequence is queried independently)
--res-embedding-folder: Directory for intermediate residue embeddings
--min-res-n: Minimum residue count for sequence filtering (default: 0)
--top-k: Number of results per query sequence (default: 100)
--threshold: Minimum similarity score (default: 0.8)
--output-csv: Export results to CSV (optional)
--use-gpu-index: Use GPU for FAISS search
--accelerator, --devices, --strategy: Inference device settings
--batch-size-res, --num-workers-res, --num-nodes-res: Residue embedding inference settings
--batch-size-aggregator, --num-workers-aggregator, --num-nodes-aggregator: Chain embedding inference settings
--log-level: Logging level (default: info)

`fm-search query db`

Compare all entries from a query database against a subject database.

fm-search query db \
  --query-db-path databases/query_set \
  --subject-db-path databases/target_set \
  --top-k 100 \
  --threshold 0.8 \
  --output-csv comparisons.csv

Key Options:

--query-db-path: Query database path
--subject-db-path: Subject database to search
--top-k: Results per query (default: 100)
--threshold: Similarity threshold (default: 0.8)
--output-csv: Export results to CSV
--use-gpu-index: Use GPU acceleration

`fm-search stats`

Display database statistics.

fm-search stats --db-path databases/my_structures

`fm-search cluster`

Cluster database embeddings using the Leiden algorithm.

fm-search cluster \
  --db-path databases/my_structures \
  --threshold 0.8 \
  --resolution 1.0 \
  --output clusters.csv \
  --max-neighbors 1000 \
  --min-cluster-size 5

Key Options:

--db-path: Database path
--threshold: Similarity threshold for edge creation (default: 0.8)
--resolution: Leiden resolution parameter - higher values create more clusters (default: 1.0)
--output: Output file (.csv or .json)
--max-neighbors: Maximum neighbors per chain (default: 1000)
--min-cluster-size: Filter out smaller clusters (optional)
--use-gpu-index: Use GPU for FAISS operations
--seed: Random seed for reproducibility (optional)

`fm-search similarity-graph`

Build a similarity graph from database embeddings and export it in GraphML format. Each node represents a chain (identified by its chain ID) and each edge carries a weight attribute with the cosine similarity score between the two connected chains.

fm-search similarity-graph \
  --db-path databases/my_structures \
  --threshold 0.8 \
  --output similarity_graph.graphml \
  --max-neighbors 1000

Key Options:

--db-path: Database path
--threshold: Minimum similarity score to create an edge (default: 0.8)
--output: Output GraphML file (default: similarity_graph.graphml)
--max-neighbors: Maximum neighbors per chain considered during k-NN search (default: 1000)
--use-gpu-index: Use GPU for FAISS operations
--log-level: Logging verbosity level (default: info)

The resulting GraphML file can be loaded directly into tools such as Gephi, Cytoscape, or Python's NetworkX:

import networkx as nx
G = nx.read_graphml("similarity_graph.graphml")

Python API

The RcsbStructureEmbedding class provides methods for computing embeddings programmatically.

Basic Usage

from foldmatch import FoldMatch

# Initialize model
model = FoldMatch(min_res=10, max_res=5000)

# Load models (optional - loads automatically on first use)
model.load_models()  # Auto-detects CUDA
# or specify device:
# import torch
# model.load_models(device=torch.device("cuda:0"))

Methods

`load_models(device=None)`

Load both residue and aggregator models.

import torch
model.load_models(device=torch.device("cuda"))

`load_residue_embedding(device=None)`

Load only the ESM3 residue embedding model.

model.load_residue_embedding()

`load_aggregator_embedding(device=None)`

Load only the aggregator model.

model.load_aggregator_embedding()

`residue_embedding(src_structure, structure_format='mmcif', chain_id=None, assembly_id=None)`

Compute per-residue embeddings for a structure.

Parameters:

src_structure: File path, URL, or file-like object
structure_format: 'mmcif', 'binarycif', or 'pdb'
chain_id: Specific chain ID (optional, uses all chains if None)
assembly_id: Assembly ID for biological assembly (optional)

Returns: torch.Tensor of shape [num_residues, embedding_dim]

# Single chain
residue_emb = model.residue_embedding(
    src_structure="1abc.cif",
    structure_format="mmcif",
    chain_id="A"
)

# All chains concatenated
all_residues = model.residue_embedding(
    src_structure="1abc.cif",
    structure_format="mmcif"
)

# Biological assembly
assembly_residues = model.residue_embedding(
    src_structure="1abc.cif",
    structure_format="mmcif",
    assembly_id="1"
)

`residue_embedding_by_chain(src_structure, structure_format='mmcif', chain_id=None)`

Compute per-residue embeddings separately for each chain.

Returns: dict[str, torch.Tensor] mapping chain IDs to embeddings

chain_embeddings = model.residue_embedding_by_chain(
    src_structure="1abc.cif",
    structure_format="mmcif"
)
# Returns: {'A': tensor(...), 'B': tensor(...), ...}

# Get specific chain
chain_a = model.residue_embedding_by_chain(
    src_structure="1abc.cif",
    chain_id="A"
)

`residue_embedding_by_assembly(src_structure, structure_format='mmcif', assembly_id=None)`

Compute residue embeddings for an assembly.

Returns: dict[str, torch.Tensor] mapping assembly ID to concatenated embeddings

assembly_emb = model.residue_embedding_by_assembly(
    src_structure="1abc.cif",
    structure_format="mmcif",
    assembly_id="1"
)
# Returns: {'1': tensor(...)}

`sequence_embedding(sequence)`

Compute residue embeddings from amino acid sequence (no structural information).

Parameters:

sequence: Amino acid sequence string (plain or FASTA format)

Returns: torch.Tensor of shape [sequence_length, embedding_dim]

# Plain sequence
seq_emb = model.sequence_embedding("ACDEFGHIKLMNPQRSTVWY")

# FASTA format
fasta = """>Protein1
ACDEFGHIKLMNPQRSTVWY
ACDEFGHIKLMNPQRSTVWY"""
seq_emb = model.sequence_embedding(fasta)

`aggregator_embedding(residue_embedding)`

Aggregate residue embeddings into a single structure-level vector.

Parameters:

residue_embedding: torch.Tensor from residue embedding methods

Returns: torch.Tensor of shape [1536]

residue_emb = model.residue_embedding("1abc.cif", chain_id="A")
structure_emb = model.aggregator_embedding(residue_emb)

`structure_embedding(src_structure, structure_format='mmcif', chain_id=None, assembly_id=None)`

End-to-end: compute residue embeddings and aggregate in one call.

# Complete structure embedding
structure_emb = model.structure_embedding(
    src_structure="1abc.cif",
    structure_format="mmcif",
    chain_id="A"
)
# Returns: tensor of shape [1536]

Complete Example

from foldmatch import FoldMatch
import torch

# Initialize
model = FoldMatch(min_res=10, max_res=5000)

# Option 1: Full structure embedding (one-shot)
embedding = model.structure_embedding(
    src_structure="1abc.cif",
    structure_format="mmcif",
    chain_id="A"
)

# Option 2: Step-by-step with residue embeddings
residue_emb = model.residue_embedding(
    src_structure="1abc.cif",
    structure_format="mmcif",
    chain_id="A"
)
structure_emb = model.aggregator_embedding(residue_emb)

# Option 3: Process multiple chains
chain_embeddings = model.residue_embedding_by_chain(
    src_structure="1abc.cif"
)
for chain_id, res_emb in chain_embeddings.items():
    chain_emb = model.aggregator_embedding(res_emb)
    print(f"Chain {chain_id}: {chain_emb.shape}")

# Sequence-based embedding
seq_emb = model.sequence_embedding("ACDEFGHIKLMNPQRSTVWY")
structure_from_seq = model.aggregator_embedding(seq_emb)

See the examples/ and tests/ directories for more use cases.

Model Architecture

The embedding model is trained to predict structural similarity by approximating TM-scores using cosine distances between embeddings. It consists of two main components:

Protein Language Model (PLM): Computes residue-level embeddings from a given 3D structure.
Residue Embedding Aggregator: A transformer-based neural network that aggregates these residue-level embeddings into a single vector.

Protein Language Model (PLM)

Residue-wise embeddings of protein structures are computed using the ESM3 generative protein language model.

Residue Embedding Aggregator

The aggregation component consists of six transformer encoder layers, each with a 3,072-neuron feedforward layer and ReLU activations. After processing through these layers, a summation pooling operation is applied, followed by 12 fully connected residual layers that refine the embeddings into a single 1,536-dimensional vector.

Testing

After installation, run the test suite:

pytest

Citation

Segura, J., et al. (2026). Multi-scale structural similarity embedding search across entire proteomes. (https://doi.org/10.1093/bioinformatics/btag058)

License

This project uses the EvolutionaryScale ESM-3 model and is distributed under the Cambrian Non-Commercial License Agreement.

Name		Name	Last commit message	Last commit date
Latest commit History 305 Commits
.github/workflows		.github/workflows
assets		assets
examples		examples
src/foldmatch		src/foldmatch
tests		tests
.dockerignore		.dockerignore
.gitignore		.gitignore
Dockerfile		Dockerfile
LICENSE.md		LICENSE.md
README.md		README.md
pyproject.toml		pyproject.toml
uv.lock		uv.lock

Folders and files

Latest commit

History

Repository files navigation

FoldMatch

Overview

Features

Installation

From PyPI

From Source (Development)

Usage

Command-Line Interface (CLI)

Structure Embedding Commands

fm-structure residue

fm-structure chain

fm-structure assembly

fm-structure download-models

Sequence Embeddings Commands

fm-sequence residue

fm-sequence chain

fm-sequence download-models

Search Commands

fm-search build structures

fm-search update structures

fm-search build embeddings

fm-search update embeddings

fm-search build sequences

fm-search update sequences

fm-search query structure

fm-search query embedding

fm-search query sequences

fm-search query db

fm-search stats

fm-search cluster

fm-search similarity-graph

Python API

Basic Usage

Methods

load_models(device=None)

load_residue_embedding(device=None)

load_aggregator_embedding(device=None)

residue_embedding(src_structure, structure_format='mmcif', chain_id=None, assembly_id=None)

residue_embedding_by_chain(src_structure, structure_format='mmcif', chain_id=None)

residue_embedding_by_assembly(src_structure, structure_format='mmcif', assembly_id=None)

sequence_embedding(sequence)

aggregator_embedding(residue_embedding)

structure_embedding(src_structure, structure_format='mmcif', chain_id=None, assembly_id=None)

Complete Example

Model Architecture

Protein Language Model (PLM)

Residue Embedding Aggregator

Testing

Citation

License

About

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Uh oh!

Contributors

Uh oh!

Languages

`fm-structure residue`

`fm-structure chain`

`fm-structure assembly`

`fm-structure download-models`

`fm-sequence residue`

`fm-sequence chain`

`fm-sequence download-models`

`fm-search build structures`

`fm-search update structures`

`fm-search build embeddings`

`fm-search update embeddings`

`fm-search build sequences`

`fm-search update sequences`

`fm-search query structure`

`fm-search query embedding`

`fm-search query sequences`

`fm-search query db`

`fm-search stats`

`fm-search cluster`

`fm-search similarity-graph`

`load_models(device=None)`

`load_residue_embedding(device=None)`

`load_aggregator_embedding(device=None)`

`residue_embedding(src_structure, structure_format='mmcif', chain_id=None, assembly_id=None)`

`residue_embedding_by_chain(src_structure, structure_format='mmcif', chain_id=None)`

`residue_embedding_by_assembly(src_structure, structure_format='mmcif', assembly_id=None)`

`sequence_embedding(sequence)`

`aggregator_embedding(residue_embedding)`

`structure_embedding(src_structure, structure_format='mmcif', chain_id=None, assembly_id=None)`

Packages