Easel¶

High-level interface to the Easel C library.

Easel is a library developed by the Eddy/Rivas Lab to facilitate the development of biological software in C. It is used by HMMER and Infernal.

Data Structures¶

Bitfield¶

class pyhmmer.easel.Bitfield¶

A statically sized sequence of booleans stored as a packed bitfield.

A bitfield is instantiated with a fixed length, and all booleans are set to False by default:

>>> bitfield = Bitfield(8)
>>> len(bitfield)
8
>>> bitfield[0]
False

Use indexing to access and edit individual bits:

>>> bitfield[0] = True
>>> bitfield[0]
True
>>> bitfield[0] = False
>>> bitfield[0]
False

__init__(length)¶: Create a new bitfield with the given length.

count(value=True)¶

Count the number occurrences of value in the bitfield.

If no argument is given, counts the number of True occurences.

Example

>>> bitfield = Bitfield(8)
>>> bitfield.count(False)
8
>>> bitfield[0] = bitfield[1] = True
>>> bitfield.count()
2

toggle(index)¶

Switch the value of one single bit.

Example

>>> bitfield = Bitfield(8)
>>> bitfield[0]
False
>>> bitfield.toggle(0)
>>> bitfield[0]
True
>>> bitfield.toggle(0)
>>> bitfield[0]
False

KeyHash¶

class pyhmmer.easel.KeyHash¶

A dynamically resized container to store byte keys using a hash table.

Internally uses Bob Jenkins’ one at a time hash, a simple and efficient hash function published in 1997 that exhibits avalanche behaviour.

Example

Add new keys to the key hash using the add method like you would with a Python set:

>>> kh = KeyHash()
>>> kh.add(b"key")
0

Check if a key hash contains a given key:

>>> b"key" in kh
True
>>> b"missing" in kh
False

Get the index associated with a key using the indexing notation:

>>> kh[b"key"]
0
>>> kh[b"missing"]
Traceback (most recent call last):
  ...
KeyError: b'missing'

Iterate over the keys of the key hash, in the order of insertion:

>>> kh.add(b"key2")
1
>>> for k in kh:
...     print(k)
b'key'
b'key2'

See also

The Wikipedia article for Bob Jenkins’ hash functions: https://en.wikipedia.org/wiki/Jenkins_hash_function

__init__()¶: Create a new empty key-hash collection.

add(item)¶

Add a new key to the hash table, and return its index.

If key was already in the hash table, the previous index is returned:

>>> kh = KeyHash()
>>> kh.add(b"first")
0
>>> kh.add(b"second")
1
>>> kh.add(b"first")
0

Parameters: key (bytes) – The key to add to the hash table.
Returns: int – The index corresponding to the added key.

New in version 0.3.0.

clear()¶: Remove all entries from the collection.

copy()¶

Create and return an exact copy of this mapping.

Example

>>> kh = KeyHash()
>>> kh.add(b"key")
0
>>> copy = kh.copy()
>>> b"key" in copy
True

Sequences¶

Sequence¶

class pyhmmer.easel.Sequence¶

An abstract biological sequence with some associated metadata.

Easel provides two different mode to store a sequence: text, or digital. In the HMMER code, changing from one mode to another mode is done in place, which allows recycling memory. However, doing so can be confusing since there is no way to know statically the representation of a sequence.

To avoid this, pyhmmer provides two subclasses of the Sequence abstract class to maintain the mode contract: TextSequence and DigitalSequence. Functions expecting sequences in digital format, like pyhmmer.hmmsearch, can then use Python type system to make sure they receive sequences in the right mode. This allows type checkers such as mypy to detect potential contract breaches at compile-time.

checksum()¶: Calculate a 32-bit checksum for the sequence.

clear()¶: Reinitialize the sequence for re-use.

copy()¶: Duplicate the sequence, and return the copy.

write(fh)¶

Write the sequence alignement to a file handle, in FASTA format.

Parameters: fh (io.IOBase) – A Python file handle, opened in binary mode.

New in version 0.3.0.

accession¶

The accession of the sequence.

Type: bytes

description¶

The description of the sequence.

Type: bytes

name¶

The name of the sequence.

Type: bytes

residue_markups¶

Extra residue markups, mapping information to each position.

Keys and values are not decoded, since they are not necessarily valid UTF-8 bytestrings.

Caution

The values of the dictionary must be the same size as the sequence itself. Trying to set a residue markup of the wrong length will raise a ValueError:

>>> seq = TextSequence(sequence="TTAATTGGT")
>>> seq.residue_markups = {b"quality": b"efcfffffcfee"}
Traceback (most recent call last):
  ...
ValueError: Residue markup annotation has an invalid length (expected 9)

New in version 0.4.6.

Type: dict

source¶

The source of the sequence, if any.

Type: bytes

taxonomy_id¶

The NCBI taxonomy ID for the source organism, if any.

New in version 0.4.6.

Type: int or None

TextSequence¶

class pyhmmer.easel.TextSequence(Sequence)¶

A biological sequence stored in text mode.

Hint

Use the sequence property to access the sequence letters as a Python string.

__init__(name=None, description=None, accession=None, sequence=None, source=None)¶: Create a new text-mode sequence with the given attributes.

copy()¶: Duplicate the text sequence, and return the copy.

digitize(alphabet)¶

Convert the text sequence to a digital sequence using alphabet.

Returns: DigitalSequence – A copy of the sequence in digital-model, digitized with alphabet.

reverse_complement()¶

Build the reverse complement of the sequence.

This method assumes that the sequence alphabet is IUPAC/DNA. If the sequence contains any unknown letters, they will be replaced by \(N\) in the reverse-complement.

Parameters: inplace (bool) – Whether or not to copy the sequence before computing its reverse complement. With False (the default), the method will return a copy of the sequence that has been reverse-complemented. With True, it will reverse-complement inplace and return None.
Raises: UserWarning – When the sequence contains unknown characters.

Example

>>> seq = TextSequence(sequence="ATGC")
>>> seq.reverse_complement().sequence
'GCAT'

Caution

The copy made when inplace is False is an exact copy, so the name, description and accession of the copy will be the same. This could lead to duplicates if you’re not careful!

New in version 0.3.0.

sequence¶

The raw sequence letters, as a Python string.

Type: str

DigitalSequence¶

class pyhmmer.easel.DigitalSequence(Sequence)¶

A biological sequence stored in digital mode.

alphabet¶

The biological alphabet used to encode this sequence to digits.

Type: Alphabet, readonly

Hint

Use the sequence property to access the sequence digits as a memory view, allowing to access the individual bytes. This can be combined with numpy.asarray to get the sequence as an array with zero-copy.

__init__(alphabet, name=None, description=None, accession=None, sequence=None, source=None)¶: Create a new digital-mode sequence with the given attributes.

New in version 0.1.4.

copy()¶: Duplicate the digital sequence, and return the copy.

reverse_complement()¶

Build the reverse complement of the sequence.

Parameters

inplace (bool) – Whether or not to copy the sequence before computing its reverse complement. With False (the default), the method will return a copy of the sequence that has been reverse-complemented. With True, it will reverse-complement inplace and return None.

Raises

ValueError – When the alphabet of the DigitalSequence does
not have a complement mapping set (e.g., Alphabet.amino) –

Caution

The copy made when inplace is False is an exact copy, so the name, description and accession of the copy will be the same. This could lead to duplicates if you’re not careful!

New in version 0.3.0.

textize()¶

Convert the digital sequence to a text sequence.

Returns: TextSequence – A copy of the sequence in text-mode.

New in version 0.1.4.

sequence¶

The raw sequence digits, as a byte vector.

Note

The internal ESL_SQ object allocates a buffer of size \(n+2\) (where \(n\) is the number of residues in the sequence), with the first and the last element of the buffer being sentinel values. This vector does not expose the sentinel values, only the \(n\) elements of the buffer in between.

Changed in version v0.4.0: Property is now a VectorU8 instead of a memoryview.

Type: VectorU8

SequenceFile¶

class pyhmmer.easel.SequenceFile¶

A wrapper around a sequence file, containing unaligned sequences.

This class supports reading sequences stored in different formats, such as FASTA, GenBank or EMBL. The format of each file can be automatically detected, but it is also possible to pass an explicit format specifier when the SequenceFile is instantiated.

New in version 0.2.0: The alphabet attribute.

__init__(file, format=None, ignore_gaps=False)¶

Create a new sequence file parser wrapping the given file.

Parameters

file (str) – The path to a file containing sequences in one of the supported file formats.
format (str, optional) – The format of the file, or None to autodetect. Supported values are: fasta, embl, genbank, ddbj, uniprot, ncbi, daemon, hmmpgmd, fmindex.
ignore_gaps (bool) – When set to True, allow ignoring gap characters (‘-‘) when they are present in ungapped formats such as fasta. With False, stick to the default Easel behaviour.

Changed in version 0.4.4: Added the ignore_gaps parameter.

close()¶: Close the file and free the resources used by the parser.

guess_alphabet()¶

Guess the alphabet of an open SequenceFile.

This method tries to guess the alphabet of a sequence file by inspecting the first sequence in the file. It returns the alphabet, or None if the file alphabet cannot be reliably guessed.

Raises

EOFError – if the file is empty.
OSError – if a parse error occurred.
ValueError – if this methods is called after the file was closed.

parse(buffer, format)¶: Parse a sequence from a binary buffer using the given format.

parseinto(seq, buffer, format)¶: Parse a sequence from a binary buffer into seq.

read(skip_info=False, skip_sequence=False)¶

Read the next sequence from the file.

Parameters

skip_info (bool) – Pass True to disable reading the sequence metadata, and only read the sequence letters. Defaults to False.
skip_sequence (bool) – Pass True to disable reading the sequence letters, and only read the sequence metadata. Defaults to False.

Returns

Sequence – The next sequence in the file, or None if all sequences were read from the file.

Raises

ValueError – When attempting to read a sequence from a closed file, or when the file could not be parsed.

Hint

This method allocates a new sequence, which is not efficient in case the sequences are being read within a tight loop. Use SequenceFile.readinto with an already initialized Sequence if you can to recycle the internal buffers.

readinto(seq, skip_info=False, skip_sequence=False)¶

Read the next sequence from the file, using seq to store data.

Parameters

seq (Sequence) – A sequence object to use to store the next entry in the file. If this sequence was used before, it must be properly reset (using the Sequence.clear method) before using it again with readinto.
skip_info (bool) – Pass True to disable reading the sequence metadata, and only read the sequence letters. Defaults to False`.
skip_sequence (bool) – Pass True to disable reading the sequence letters, and only read the sequence metadata. Defaults to False.

Returns

Sequence – A reference to seq that was passed as an input, or None if no sequences are left in the file.

Raises

ValueError – When attempting to read a sequence from a closed file, or when the file could not be parsed.

Example

Use SequenceFile.readinto to loop over the sequences in a file while recycling the same Sequence buffer:

>>> with SequenceFile("vendor/hmmer/testsuite/ecori.fa") as sf:
...     seq = TextSequence()
...     while sf.readinto(seq) is not None:
...         # ... process seq here ... #
...         seq.clear()

set_digital(alphabet)¶

Set the SequenceFile to read in digital mode with alphabet.

This method can be called even after the first sequences have been read; it only affects subsequent sequences in the file.

Returns: Alphabet – The alphabet it was given. Useful to wrap guess_alphabet calls and get the resulting alphabet.

Changed in version 0.4.0: Returns the Alphabet given as argument instead of None.

Alignments¶

MSA¶

class pyhmmer.easel.MSA¶

An abstract alignment of multiple sequences.

Hint

Use len(msa) to get the number of columns in the alignment, and len(msa.sequences) to get the number of sequences (i.e. the number of rows).

checksum()¶: Calculate a 32-bit checksum for the multiple sequence alignment.

write(fh, format)¶

Write the multiple sequence alignement to a file handle.

Parameters

fh (io.IOBase) – A Python file handle, opened in binary mode.
format (str) – The name of the multiple sequence alignment file format to use.

New in version 0.3.0.

accession¶

The accession of the alignment, if any.

Type: bytes or None

author¶

The author of the alignment, if any.

Type: bytes or None

description¶

The description of the sequence, if any.

Type: bytes or None

name¶

The name of the alignment, if any.

Type: bytes or None

TextMSA¶

class pyhmmer.easel.TextMSA(MSA)¶

A multiple sequence alignement stored in text mode.

__init__(name=None, description=None, accession=None, sequences=None, author=None)¶

Create a new text-mode alignment with the given sequences.

Parameters

name (bytes, optional) – The name of the alignment, if any.
description (bytes, optional) – The description of the alignment, if any.
accession (bytes, optional) – The accession of the alignment, if any.
sequences (collection of TextSequence) – The sequences to store in the multiple sequence alignment. All sequences must have the same length. They also need to have distinct names.
author (bytes, optional) – The author of the alignment, often used to record the aligner it was created with.

Raises

ValueError – When the alignment cannot be created from the given sequences.
TypeError – When sequences is not an iterable of TextSequence objects.

Example

>>> s1 = TextSequence(name=b"seq1", sequence="ATGC")
>>> s2 = TextSequence(name=b"seq2", sequence="ATGC")
>>> msa = TextMSA(name=b"msa", sequences=[s1, s2])
>>> len(msa)
4

Changed in version 0.3.0: Allow creating an alignment from an iterable of TextSequence.

copy()¶: Duplicate the text sequence alignment, and return the copy.

digitize(alphabet)¶

Convert the text alignment to a digital alignment using alphabet.

Returns: DigitalMSA – An alignment in digital mode containing the same sequences digitized with alphabet.

sequences¶

A view of the sequences in the alignment.

This property lets you access the individual sequences in the multiple sequence alignment as TextSequence instances.

Example

Query the number of sequences in the alignment with len, or access individual members via indexing notation:

>>> s1 = TextSequence(name=b"seq1", sequence="ATGC")
>>> s2 = TextSequence(name=b"seq2", sequence="ATGC")
>>> msa = TextMSA(name=b"msa", sequences=[s1, s2])
>>> len(msa.sequences)
2
>>> msa.sequences[0].name
b'seq1'

Caution

Sequences in the list are copies, so editing their attributes will have no effect on the alignment:

>>> msa.sequences[0].name
b'seq1'
>>> msa.sequences[0].name = b"seq1bis"
>>> msa.sequences[0].name
b'seq1'

Support for this feature will be added in a future version, but can be circumvented for now by forcingly setting the updated version of the object:

>>> seq = msa.sequences[0]
>>> seq.name = b"seq1bis"
>>> msa.sequences[0] = seq
>>> msa.sequences[0].name
b'seq1bis'

New in version 0.3.0.

Type: _TextMSASequences

DigitalMSA¶

class pyhmmer.easel.DigitalMSA(MSA)¶

A multiple sequence alignment stored in digital mode.

alphabet¶

The biological alphabet used to encode this sequence alignment to digits.

Type: Alphabet

__init__(alphabet, name=None, description=None, accession=None, sequences=None, author=None)¶

Create a new digital-mode alignment with the given sequences.

Parameters

alphabet (Alphabet) – The alphabet of the alignmed sequences.
name (bytes, optional) – The name of the alignment, if any.
description (bytes, optional) – The description of the alignment, if any.
accession (bytes, optional) – The accession of the alignment, if any.
sequences (iterable of DigitalSequence) – The sequences to store in the multiple sequence alignment. All sequences must have the same length and alphabet. They also need to have distinct names set.
author (bytes, optional) – The author of the alignment, often used to record the aligner it was created with.

Changed in version 0.3.0: Allow creating an alignment from an iterable of DigitalSequence.

copy()¶: Duplicate the digital sequence alignment, and return the copy.

textize()¶

Convert the digital alignment to a text alignment.

Returns: TextMSA – A copy of the alignment in text-mode.

New in version 0.3.0.

sequences¶

A view of the sequences in the alignment.

This property lets you access the individual sequences in the multiple sequence alignment as DigitalSequence instances.

See also

The documentation for the TextMSA.sequences property, which contains some additional information.

New in version 0.3.0.

Type: _DigitalMSASequences

MSAFile¶

class pyhmmer.easel.MSAFile¶

A wrapper around a multiple-alignment file.

This class supports reading sequences stored in different formats, such as Stockholm, A2M, PSI-BLAST or Clustal.

__init__(*args, **kwargs)¶: Initialize self. See help(type(self)) for accurate signature.

close()¶: Close the file and free the resources used by the parser.

guess_alphabet()¶

Guess the alphabet of an open MSAFile.

This method tries to guess the alphabet of a multiple-alignment file by inspecting the first entry in the file. It returns the alphabet, or None if the file alphabet cannot be reliably guessed.

Raises

EOFError – if the file is empty.
OSError – if a parse error occurred.
ValueError – if this methods is called after the file was closed.

read()¶

Read the next alignment from the file.

Returns: MSA – The next alignment in the file, or None if all the alignments were read from the file already.
Raises: ValueError – When attempting to read an alignment from a closed file, or when the file could not be parsed.

Hint

This method allocates a new alignment, which is not efficient in case the sequences are being read within a tight loop. Use SequenceFile.readinto with an already initialized Sequence if you can to recycle the internal buffers.

set_digital(alphabet)¶

Set the MSAFile to read in digital mode with alphabet.

This method can be called even after the first alignment have been read; it only affects subsequent sequences in the file.

Returns: Alphabet – The alphabet it was given. Useful to wrap guess_alphabet calls and get the resulting alphabet.

Changed in version 0.4.0: Returns the Alphabet given as argument instead of None.

Linear Algebra¶

Vector¶

class pyhmmer.easel.Vector¶

An abstract 1D array of fixed size.

New in version 0.4.0.

argmax()¶

Return index of the maximum element in the vector.

Raises: ValueError – When called on an empty vector.

argmin()¶

Return index of the minimum element in the vector.

Raises: ValueError – When called on an empty vector.

copy()¶: Create a copy of the vector, allocating a new buffer.

max()¶

Return value of the maximum element in the vector.

Raises: ValueError – When called on an empty vector.

min()¶

Return value of the minimum element in the vector.

Raises: ValueError – When called on an empty vector.

reverse()¶: Reverse the vector, in place.

sum()¶: Returns the scalar sum of all elements in the vector.

zeros(n)¶: Create a vector of size n filled with zeros.

format¶

The format of each item in the vector.

See also

The array module of the Python standard library for a detail about available type codes.

New in version 0.4.6.

Type: str

itemsize¶

The size of each item in the vector, in bytes.

New in version 0.4.6.

Type: int

shape¶

The shape of the vector.

Type: tuple

strides¶

The strides of the vector.

Type: tuple

VectorF¶

class pyhmmer.easel.VectorF¶

A vector storing single-precision floating point numbers.

Individual elements of a vector can be accessed and modified with the usual indexing notation:

>>> v = VectorF([1.0, 2.0, 3.0])
>>> v[0]
1.0
>>> v[-1]
3.0
>>> v[0] = v[-1] = 4.0
>>> v
VectorF([4.0, 2.0, 4.0])

Slices are also supported, and they do not copy data (use the copy method to allocate a new vector):

>>> v = VectorF(range(10))
>>> v[2:5]
VectorF([2.0, 3.0, 4.0])
>>> v[5:-1] = 10.0
>>> v
VectorF([0.0, 1.0, 2.0, 3.0, 4.0, 10.0, 10.0, 10.0, 10.0, 9.0])

Addition and multiplication is supported for scalars, in place or not:

>>> v = VectorF([1.0, 2.0, 3.0])
>>> v += 1
>>> v
VectorF([2.0, 3.0, 4.0])
>>> v * 3
VectorF([6.0, 9.0, 12.0])

Pairwise operations can also be performed, but only on vectors of the same dimension and precision:

>>> v = VectorF([1.0, 2.0, 3.0])
>>> v * v
VectorF([1.0, 4.0, 9.0])
>>> v += VectorF([3.0, 4.0, 5.0])
>>> v
VectorF([4.0, 6.0, 8.0])
>>> v *= VectorF([1.0])
Traceback (most recent call last):
  ...
ValueError: cannot pairwise multiply vectors of different size

Objects of this type support the buffer protocol, and can be viewed as a numpy.ndarray of one dimension using the numpy.asarray function, and can be passed without copy to most numpy functions:

>>> v = VectorF([1.0, 2.0, 3.0])
>>> numpy.asarray(v)
array([1., 2., 3.], dtype=float32)
>>> numpy.log2(v)
array([0.       , 1.       , 1.5849625], dtype=float32)

New in version 0.4.0.

__init__(*args, **kwargs)¶: Initialize self. See help(type(self)) for accurate signature.

argmax()¶

Return index of the maximum element in the vector.

Raises: ValueError – When called on an empty vector.

argmin()¶

Return index of the minimum element in the vector.

Raises: ValueError – When called on an empty vector.

copy()¶: Create a copy of the vector, allocating a new buffer.

max()¶

Return value of the maximum element in the vector.

Raises: ValueError – When called on an empty vector.

min()¶

Return value of the minimum element in the vector.

Raises: ValueError – When called on an empty vector.

normalize()¶: Normalize a vector so that all elements sum to 1.

Caution

If sum is zero, sets all elements to \(\frac{1}{n}\), where \(n\) is the size of the vector.

reverse()¶: Reverse the vector, in place.

sum()¶

Returns the scalar sum of all elements in the vector.

Floating point summations use Kahan compensated summation, in order to minimize roundoff error accumulation. Additionally, they are most accurate if the vector is sorted in increasing order, from small to large, so you may consider sorting the vector before summing it.

format¶

itemsize¶

VectorU8¶

class pyhmmer.easel.VectorU8¶

A vector storing byte-sized unsigned integers.

New in version v0.4.0.

__init__(*args, **kwargs)¶: Initialize self. See help(type(self)) for accurate signature.

argmax()¶

Return index of the maximum element in the vector.

Raises: ValueError – When called on an empty vector.

argmin()¶

Return index of the minimum element in the vector.

Raises: ValueError – When called on an empty vector.

copy()¶: Create a copy of the vector, allocating a new buffer.

max()¶

Return value of the maximum element in the vector.

Raises: ValueError – When called on an empty vector.

min()¶

Return value of the minimum element in the vector.

Raises: ValueError – When called on an empty vector.

reverse()¶: Reverse the vector, in place.

sum()¶

Returns the scalar sum of all elements in the vector.

Caution

The sum is wrapping:

>>> vec = VectorU8([255, 2])
>>> vec.sum()
1

format¶

itemsize¶

Matrix¶

class pyhmmer.easel.Matrix¶

An abstract 2D array of fixed size.

New in version 0.4.0.

argmax()¶

Return the coordinates of the maximum element in the matrix.

Raises: ValueError – When called on an empty matrix.

argmin()¶

Return the coordinates of the minimum element in the matrix.

Raises: ValueError – When called on an empty matrix.

copy()¶: Create a copy of the matrix, allocating a new buffer.

max()¶

Return the value of the maximum element in the matrix.

Raises: ValueError – When called on an empty matrix.

min()¶

Return the value of the minimum element in the matrix.

Raises: ValueError – When called on an empty matrix.

sum()¶: Return the sum of all elements in the matrix.

zeros(m, n)¶: Create a new \(m \times n\) matrix filled with zeros.

format¶

The format of each item in the matrix.

See also

The array module of the Python standard library for a detail about available type codes.

New in version 0.4.7.

Type: str

itemsize¶

The size of each item in the matrix, in bytes.

New in version 0.4.7.

Type: int

shape¶

The shape of the matrix.

Example

>>> m = MatrixF([ [1.0, 2.0], [3.0, 4.0], [5.0, 6.0] ])
>>> m.shape
(3, 2)

Type: tuple

strides¶

The strides of the matrix.

Type: tuple

MatrixF¶

class pyhmmer.easel.MatrixF¶

A matrix storing single-precision floating point numbers.

Use indexing notation to access and edit individual elements of the matrix:

>>> m = MatrixF.zeros(2, 2)
>>> m[0, 0] = 3.0
>>> m
MatrixF([[3.0, 0.0], [0.0, 0.0]])

Indexing can also be performed at the row-level to get a VectorF without copying the underlying data:

>>> m = MatrixF([ [1.0, 2.0], [3.0, 4.0] ])
>>> m[0]
VectorF([1.0, 2.0])

Objects of this type support the buffer protocol, and can be viewed as a numpy.ndarray with two dimensions using the numpy.asarray function, and can be passed without copy to most numpy functions:

>>> m = MatrixF([ [1.0, 2.0], [3.0, 4.0] ])
>>> numpy.asarray(m)
array([[1., 2.],
       [3., 4.]], dtype=float32)
>>> numpy.log2(m)
array([[0.       , 1.       ],
       [1.5849625, 2.       ]], dtype=float32)

New in version 0.4.0.

__init__(*args, **kwargs)¶: Initialize self. See help(type(self)) for accurate signature.

argmax()¶

Return the coordinates of the maximum element in the matrix.

Raises: ValueError – When called on an empty matrix.

argmin()¶

Return the coordinates of the minimum element in the matrix.

Raises: ValueError – When called on an empty matrix.

copy()¶: Create a copy of the matrix, allocating a new buffer.

max()¶

Return the value of the maximum element in the matrix.

Raises: ValueError – When called on an empty matrix.

min()¶

Return the value of the minimum element in the matrix.

Raises: ValueError – When called on an empty matrix.

sum()¶: Return the sum of all elements in the matrix.

format¶

itemsize¶

MatrixU8¶

class pyhmmer.easel.MatrixU8¶

A matrix storing byte-sized unsigned integers.

New in version v0.4.0.

__init__(*args, **kwargs)¶: Initialize self. See help(type(self)) for accurate signature.

argmax()¶

Return the coordinates of the maximum element in the matrix.

Raises: ValueError – When called on an empty matrix.

argmin()¶

Return the coordinates of the minimum element in the matrix.

Raises: ValueError – When called on an empty matrix.

copy()¶: Create a copy of the matrix, allocating a new buffer.

max()¶

Return the value of the maximum element in the matrix.

Raises: ValueError – When called on an empty matrix.

min()¶

Return the value of the minimum element in the matrix.

Raises: ValueError – When called on an empty matrix.

sum()¶: Return the sum of all elements in the matrix.

format¶

itemsize¶

Miscellaneous¶

Alphabet¶

class pyhmmer.easel.Alphabet¶

A biological alphabet, including additional marker symbols.

This type is used to share an alphabet to several objects in the easel and plan7 modules. Reference counting helps sharing the same instance everywhere, instead of reallocating memory every time an alphabet is needed.

Use the factory class methods to obtain a default Alphabet for one of the three standard biological alphabets:

>>> dna = Alphabet.dna()
>>> rna = Alphabet.rna()
>>> aa  = Alphabet.amino()

amino()¶: Create a default amino-acid alphabet.

dna()¶: Create a default DNA alphabet.

rna()¶: Create a default RNA alphabet.

K¶

The alphabet size, counting only actual alphabet symbols.

Example

>>> Alphabet.dna().K
4
>>> Alphabet.amino().K
20

Type: int

Kp¶

The complete alphabet size, including marker symbols.

Example

>>> Alphabet.dna().Kp
18
>>> Alphabet.amino().Kp
29

Type: int

symbols¶

The symbols composing the alphabet.

Example

>>> Alphabet.dna().symbols
'ACGT-RYMKSWHBVDN*~'
>>> Alphabet.rna().symbols
'ACGU-RYMKSWHBVDN*~'

Type: str

Randomness¶

class pyhmmer.easel.Randomness¶

A portable, thread-safe random number generator.

Methods with an implementation in Easel are named after the equivalent methods of random.Random.

New in version 0.4.2.

__init__(seed=None, fast=False)¶

Create a new random number generator with the given seed.

Parameters

seed (int) – The seed to initialize the generator with. If 0 or None is given, an arbitrary seed will be chosen using the system clock.
fast (bool) – If True, use a linear congruential generator (LCG), which is low quality and should only be used for integration with legacy code. With False, use the Mersenne Twister MT19937 algorithm instead.

copy()¶: Return a copy of the random number generator in the same exact state.

getstate()¶: Get a tuple containing the current state.

is_fast()¶: Returns whether or not the linear congruential generator is in use.

normalvariate(mu, sigma)¶

Generate a Gaussian-distributed sample.

Parameters

mu (float) – The mean of the Gaussian being sampled.
sigma (float) – The standard deviation of the Gaussian being sampled.

random()¶: Generate a uniform random deviate on \(\left[ 0, 1 \right)\).

seed()¶

Reinitialize the random number generator with the given seed.

Parameters: n (int, optional) – The seed to use. If 0 or None, an arbitrary seed will be chosen using the current time.

setstate(state)¶: Restores the state of the random number generator.

Sequence / Subsequence Index¶

class pyhmmer.easel.SSIReader¶

A read-only handler for sequence/subsequence index file.

class Entry(fd, record_offset, data_offset, record_length)¶

property data_offset¶: Alias for field number 2

property fd¶: Alias for field number 0

property record_length¶: Alias for field number 3

property record_offset¶: Alias for field number 1

class FileInfo(name, format)¶

property format¶: Alias for field number 1

property name¶: Alias for field number 0

__init__(file)¶

Create a new SSI file reader for the file at the given location.

Parameters: file (str) – The path to a sequence/subsequence index file to read.

close()¶: Close the SSI file reader.

file_info(fd)¶: Retrieve the FileInfo of the descriptor.

find_name(key)¶: Retrieve the Entry for the given name.

class pyhmmer.easel.SSIWriter¶

A writer for sequence/subsequence index files.

__init__(file)¶

Create a new SSI file write for the file at the given location.

Parameters

file (str) – The path to a sequence/subsequence index file to write.
exclusive (bool) – Whether or not to create a file if one does not exist.

Raises

FileNotFoundError – When the path to the file cannot be resolved.
FileExistsError – When the file exists and exclusive is True.

add_alias(alias, key)¶: Make alias an alias of key in the index.

add_file(filename, format=0)¶

Add a new file to the index.

Parameters

filename (str) – The name of the file to register.
format (int) – A format code to associate with the file, or 0.

Returns

int – The filehandle associated with the new indexed file.

add_key(key, fd, record_offset, data_offset=0, record_length=0)¶: Add a new entry to the index with the given key.

close()¶: Close the SSI file writer.