biotite.structure.AtomArray¶

class biotite.structure.AtomArray(length)[source]¶

Bases: _AtomArrayBase

An array representation of a model consisting of multiple atoms.

An AtomArray can be seen as a list of Atom instances. Instead of using directly a list, this class uses an NumPy ndarray for each annotation category and the coordinates. These coordinates can be accessed directly via the coord attribute. The annotations are accessed either via the category as attribute name or the get_annotation(), set_annotation() method. Usage of custom annotations is achieved via add_annotation() or set_annotation(). A detailed description of each annotation category can be viewed here.

In order to get an an subarray of an AtomArray, NumPy style indexing is used. This includes slices, boolean arrays, index arrays and even Ellipsis notation. Using a single integer as index returns a single Atom instance.

Inserting or appending an AtomArray to another AtomArray is done with the ‘+’ operator. Only the annotation categories, which are existing in both arrays, are transferred to the new array.

Optionally, an AtomArray can store chemical bond information via a BondList object. It can be accessed using the bonds attribute. If no bond information is available, bonds is None. Consequently the bond information can be removed from the AtomArray, by setting bonds to None. When indexing the AtomArray the atom indices in the associated BondList are updated as well, hence the indices in the BondList will always point to the same atoms. If two AtomArray instances are concatenated, the resulting AtomArray will contain the merged BondList if at least one of the operands contains bond information.

The box attribute contains the box vectors of the unit cell or the MD simulation box, respectively. Hence, it is a 3 x 3 ndarray with the vectors in the last dimension. If no box is provided, the attribute is None. Setting the box attribute to None means removing the box from the atom array.

Parameters

lengthint: The fixed amount of atoms in the array.

Examples

Creating an atom array from atoms:

>>> atom1 = Atom([1,2,3], chain_id="A")
>>> atom2 = Atom([2,3,4], chain_id="A")
>>> atom3 = Atom([3,4,5], chain_id="B")
>>> atom_array = array([atom1, atom2, atom3])
>>> print(atom_array.array_length())
3

Accessing an annotation array:

>>> print(atom_array.chain_id)
['A' 'A' 'B']

Accessing the coordinates:

>>> print(atom_array.coord)
[[1. 2. 3.]
 [2. 3. 4.]
 [3. 4. 5.]]

NumPy style filtering:

>>> atom_array = atom_array[atom_array.chain_id == "A"]
>>> print(atom_array.array_length())
2

Inserting an atom array:

>>> insert = array([Atom([7,8,9], chain_id="C")])
>>> atom_array = atom_array[0:1] + insert + atom_array[1:2]
>>> print(atom_array.chain_id)
['A' 'C' 'A']

Attributes

{annot}ndarray: Multiple n-length annotation arrays.
coordndarray, dtype=float, shape=(n,3): ndarray containing the x, y and z coordinate of the atoms.
bondsBondList or None: A BondList, specifying the indices of atoms that form a chemical bond.
boxndarray, dtype=float, shape=(3,3) or None: The surrounding box. May represent a MD simulation box or a crystallographic unit cell.
shapetuple of int: Tuple of array dimensions.

add_annotation(category, dtype)¶

Add an annotation category, if not already existing.

Initially the new annotation is filled with the zero representation of the given type.

Parameters

categorystr: The annotation category to be added.
dtypetype or str: A type instance or a valid NumPy dtype string. Defines the type of the annotation