skbio.metadata.IntervalMetadata¶

class skbio.metadata.IntervalMetadata(upper_bound, copy_from=None)[source]¶

Stores the interval features.

IntervalMetadata object allows storage, modification, and querying of interval features covering a region of a single coordinate system. For instance, this can be used to store functional annotations about genes across a genome. This object is also applied to the sequence alignment.

This object is typically coupled with another object, such as a Sequence object (or its child class), or a TabularMSA object.

Parameters:

upper_bound (int or None) – Defines the exclusive upper bound of the interval features. No coordinate can be greater than it. None means that the coordinate space is unbounded. copy_from (IntervalMetadata or None, optional) – Create a new object from the input IntervalMetadata object by shallow copying if it is not None. The upper bound of the new object will be updated with the upper_bound parameter specified.

Notes

This class stores coordinates of all feature bounds into a interval tree. It allows the speed up of query-by-bound. The building of interval tree is deferred until necessary to save computation. It is updated from all coordinates only when you need to fetch info from the interval tree.

When you add a method into this class and if you method need to fetch info from IntervalMetadata._interval_tree, you should decorate it with _rebuild_tree. This decorator will check if the current interval tree is stale and will update it if so. Additionally, if your method add, delete, or changes the coordinates of any interval features, you should set self._is_stale_tree to True at the end of your method to indicate the interval tree becomes stale.

See also

skbio.metadata.Interval

Examples

Let’s say we have a sequence of length 10 and want to add annotation to it. Create an IntervalMetadata object:

>>> from skbio.metadata import Interval, IntervalMetadata
>>> im = IntervalMetadata(10)

Let’s add annotations of 3 genes:

>>> im.add(bounds=[(3, 9)],
...        metadata={'gene': 'sagB'})  # doctest: +ELLIPSIS
Interval(interval_metadata=..., bounds=[(3, 9)], fuzzy=[(False, False)], metadata={'gene': 'sagB'})
>>> im.add(bounds=[(3, 7)],
...        metadata={'gene': 'sagC'})  # doctest: +ELLIPSIS
Interval(interval_metadata=..., bounds=[(3, 7)], fuzzy=[(False, False)], metadata={'gene': 'sagC'})
>>> im.add(bounds=[(1, 2), (4, 7)],
...        metadata={'gene': 'sagA'})  # doctest: +ELLIPSIS
Interval(interval_metadata=..., bounds=[(1, 2), (4, 7)], fuzzy=[(False, False), (False, False)], metadata={'gene': 'sagA'})

Show the object representation:

>>> im    # doctest: +ELLIPSIS
3 interval features
-------------------
Interval(interval_metadata=..., bounds=[(3, 9)], fuzzy=[(False, False)], metadata={'gene': 'sagB'})
Interval(interval_metadata=..., bounds=[(3, 7)], fuzzy=[(False, False)], metadata={'gene': 'sagC'})
Interval(interval_metadata=..., bounds=[(1, 2), (4, 7)], fuzzy=[(False, False), (False, False)], metadata={'gene': 'sagA'})

We can sort the genes by their bounds:

>>> im.sort()
>>> im    # doctest: +ELLIPSIS
3 interval features
-------------------
Interval(interval_metadata=..., bounds=[(1, 2), (4, 7)], fuzzy=[(False, False), (False, False)], metadata={'gene': 'sagA'})
Interval(interval_metadata=..., bounds=[(3, 7)], fuzzy=[(False, False)], metadata={'gene': 'sagC'})
Interval(interval_metadata=..., bounds=[(3, 9)], fuzzy=[(False, False)], metadata={'gene': 'sagB'})

Query the genes by bound and/or metadata:

>>> intvls = im.query([(1, 2)], metadata={'gene': 'foo'})
>>> list(intvls)
[]
>>> intvls = im.query([(7, 9)])
>>> list(intvls)  # doctest: +ELLIPSIS
[Interval(interval_metadata=..., bounds=[(3, 9)], fuzzy=[(False, False)], metadata={'gene': 'sagB'})]
>>> intvls = im.query(metadata={'gene': 'sagA'})
>>> intvls = list(intvls)
>>> intvls  # doctest: +ELLIPSIS
[Interval(interval_metadata=..., bounds=[(1, 2), (4, 7)], fuzzy=[(False, False), (False, False)], metadata={'gene': 'sagA'})]

Drop the gene(s) we get from query:

>>> im.drop(intvls)
>>> im.sort()
>>> im   # doctest: +ELLIPSIS
2 interval features
-------------------
Interval(interval_metadata=..., bounds=[(3, 7)], fuzzy=[(False, False)], metadata={'gene': 'sagC'})
Interval(interval_metadata=..., bounds=[(3, 9)], fuzzy=[(False, False)], metadata={'gene': 'sagB'})

Attributes

default_write_format
lower_bound	The inclusive lower bound of interval features.
num_interval_features	The total number of interval features.
upper_bound	The exclusive upper bound of interval features.

Built-ins

copy.copy(im)	Return a shallow copy.
copy.deepcopy(im)	Return a deep copy.
im1 == im2	Test if this object is equal to another.
__init_subclass__	This method is called when a class is subclassed.
im1 != im2	Test if this object is not equal to another.

Methods

add(bounds[, fuzzy, metadata])	Create and add an Interval to this IntervalMetadata.
concat(interval_metadata)	Concatenate an iterable of IntervalMetadata objects.
drop(intervals[, negate])	Drops Interval objects.
merge(other)	Merge the interval features of another IntervalMetadata object.
query([bounds, metadata])	Yield Interval object with the bounds and attributes.
read(file[, format])	Create a new IntervalMetadata instance from a file.
sort([ascending])	Sort interval features by their coordinates.
write(file[, format])	Write an instance of IntervalMetadata to a file.