proteindf_bridge.bond module

class proteindf_bridge.bond.Bond

Bases: object

Heuristic bond detector based on covalent radii.

Bonds are established between pairs of atoms (p, q) satisfying:

distance(p, q) <= covalent_radius(p) + covalent_radius(q) + 0.45 Å

where 0.45 Å is the covalent bond tolerance.

Key behaviors:

• No bonds across models: When the structure contains multiple models (satisfying Format.is_models), bonds are not formed between atoms belonging to different models.

• Preservation of existing bonds: Pre-existing bonds and their bond orders in the AtomGroup are preserved and not duplicated.

• Dense matrix threshold: If the number of atoms exceeds 2000, the pairwise matrices (distmat and bondmat) are None to prevent excessive memory usage, while bond connectivity is still recorded in the AtomGroup in O(N) time.

bondmat

Pairwise bond connectivity matrix, or None if atom count exceeds 2000.

Returns a SymmetricMatrix (1.0 for bonded, 0.0 otherwise) when the number of atoms is 2000 or fewer, or None if there are more than 2000 atoms to protect against O(N^2) memory consumption.

distmat

Pairwise atomic distance matrix, or None if atom count exceeds 2000.

Returns a SymmetricMatrix when the number of atoms is 2000 or fewer, or None if there are more than 2000 atoms to protect against O(N^2) memory consumption.

setup(mol)

Detects and sets up covalent bonds for the given molecular structure.

Computes interatomic distances and adds detected covalent bonds to mol. Existing bonds and their bond orders are preserved without duplicate registration. If mol represents multiple models, no bonds are formed across different models.

Parameters:

mol (AtomGroup) – Molecular structure to analyze and update.

Returns:

None

Raises:

• TypeError – If mol is not an AtomGroup instance.

• BrValueError – If an atom lacks a covalent radius (e.g., dummy atom). In this case, mol is not modified.