Foldcomp

Instead of reading from or writing to a file, you can manually initialize the mtar_t struct to work with custom streams.

Steps to use a custom stream:

1. Manually initialize the mtar_t struct.
2. Set the stream pointer.
3. Provide the required callback functions.
4. Ensure all unused fields are zero-initialized.
5. Do not call mtar_open if you have initialized the struct manually.

Required Callbacks for Reading

Required Callbacks for Writing

All callbacks must return MTAR_ESUCCESS on success or a negative integer on error.

Foldcomp provides a Python helper to download large prebuilt databases of predicted protein structures. Use foldcomp.setup('database_name') to initiate the download.

Available Databases

• ESMAtlas:
  • Full: foldcomp.setup('esmatlas')
  • v2023_02: foldcomp.setup('esmatlas_v2023_02')
  • High-quality: foldcomp.setup('highquality_clust30')
• AlphaFoldDB Uniprot:
  • v6: foldcomp.setup('afdb_uniprot_v6')
  • v4: foldcomp.setup('afdb_uniprot_v4')
• AlphaFoldDB Swiss-Prot:
  • foldcomp.setup('afdb_swissprot_v4')
• AlphaFoldDB Model Organisms:
  • foldcomp.setup('h_sapiens')
  • foldcomp.setup('a_thaliana'), foldcomp.setup('c_albicans'), foldcomp.setup('c_elegans'), foldcomp.setup('d_discoideum'), foldcomp.setup('d_melanogaster'), foldcomp.setup('d_rerio'), foldcomp.setup('e_coli'), foldcomp.setup('g_max'), foldcomp.setup('m_jannaschii'), foldcomp.setup('m_musculus'), foldcomp.setup('o_sativa'), foldcomp.setup('r_norvegicus'), foldcomp.setup('s_cerevisiae'), foldcomp.setup('s_pombe'), foldcomp.setup('z_mays')
• AlphaFoldDB Cluster Representatives:
  • foldcomp.setup('afdb_rep_v4')
  • foldcomp.setup('afdb_rep_dark_v4')

import foldcomp; foldcomp.setup('afdb_swissprot_v4')

The foldcomp executable provides a command-line interface for lossy compression of protein structures. It encodes torsion angles to save space and can reconstruct 3D coordinates from the encoded data.

Basic Usage

• Compress a single file: foldcomp compress <input.pdb|input.cif> <output.fcz>
• Decompress a file: foldcomp decompress <input.fcz> <output.pdb>

Batch Operations

Foldcomp supports processing directories, tar files, and databases:

• Compress a directory/tar: foldcomp compress [-t threads] <dir|tar(.gz)> <output_dir|tar|db>
• Decompress a directory/tar: foldcomp decompress [-t threads] <dir|tar(.gz)|db> <output_dir|tar>

Extraction and Checking

• Extract metadata: Use extract to pull pLDDT scores or amino acid sequences from .fcz files.
• Check integrity: Use check to verify the validity of .fcz files.

# Compress a PDB file
foldcomp compress input.pdb output.fcz

# Decompress an FCZ file
foldcomp decompress output.fcz output.pdb

# Compress a directory using 4 threads
foldcomp compress -t 4 ./pdb_dir ./output_dir

The Foldcomp Python API allows for programmatic decompression and data extraction from compressed FCZ files and databases.

Decompressing a single FCZ file

Use foldcomp.decompress(fcz_binary) to get the filename and the PDB content as a string.

Extracting data from an FCZ file

Use foldcomp.get_data(fcz_binary) (or foldcomp.get_data(pdb_path)) to retrieve a dictionary containing:

• phi: phi angles (C-N-CA-C)
• psi: psi angles (N-CA-C-N)
• omega: omega angles (CA-C-N-CA)
• torsion_angles: backbone torsion angles (phi + psi + omega)
• bond_angles: backbone bond angles
• residues: amino acid residues as a string
• coordinates: backbone coordinates

Iterating over a database

Use foldcomp.open(path, ids=[...]) to iterate over specific entries in a Foldcomp database.

import foldcomp

# Decompressing a single FCZ file
with open("test/compressed.fcz", "rb") as fcz:
    fcz_binary = fcz.read()
    (name, pdb) = foldcomp.decompress(fcz_binary)
    with open(name, "w") as pdb_file:
        pdb_file.write(pdb)

# Getting data as a dictionary
data_dict = foldcomp.get_data(fcz_binary)
print(data_dict["phi"])

# Iterating over a database
ids = ["d1asha_", "d1it2a_"]
with foldcomp.open("test/example_db", ids=ids) as db:
    for (name, pdb) in db:
        with open(name + ".pdb", "w") as pdb_file:
            pdb_file.write(pdb)