biotite 1.5.0__cp312-cp312-manylinux_2_24_x86_64.manylinux_2_28_x86_64.whl

biotite/sequence/align/selector.pyx

@@ -0,0 +1,954 @@
+# This source code is part of the Biotite package and is distributed
+# under the 3-Clause BSD License. Please see 'LICENSE.rst' for further
+# information.
+
+__name__ = "biotite.sequence.align"
+__author__ = "Patrick Kunzmann"
+__all__ = ["MinimizerSelector", "SyncmerSelector", "CachedSyncmerSelector",
+           "MincodeSelector"]
+
+cimport cython
+cimport numpy as np
+
+import numpy as np
+from .kmeralphabet import KmerAlphabet
+
+
+ctypedef np.int64_t int64
+ctypedef np.uint32_t uint32
+
+
+# Obtained from 'np.iinfo(np.int64).max'
+cdef int64 MAX_INT_64 = 9223372036854775807
+
+
+class MinimizerSelector:
+    """
+    MinimizerSelector(kmer_alphabet, window, permutation=None)
+
+    Selects the *minimizers* in sequences.
+
+    In a rolling window of *k-mers*, the minimizer is defined as the
+    *k-mer* with the minimum *k-mer* code :footcite:`Roberts2004`.
+    If the same minimum *k-mer* appears twice in a window, the leftmost
+    *k-mer* is selected as minimizer.
+
+    Parameters
+    ----------
+    kmer_alphabet : KmerAlphabet
+        The *k-mer* alphabet that defines the *k-mer* size and the type
+        of sequence this :class:`MinimizerSelector` can be applied on.
+    window : int
+        The size of the rolling window, where the minimizers are
+        searched in.
+        In other words this is the number of *k-mers* per window.
+        The window size must be at least 2.
+    permutation : Permutation
+        If set, the *k-mer* order is permuted, i.e.
+        the minimizer is chosen based on the ordering of the sort keys
+        from :class:`Permutation.permute()`.
+        By default, the standard order of the :class:`KmerAlphabet` is
+        used.
+        This standard order is often the lexicographical order, which is
+        known to yield suboptimal *density* in many cases
+        :footcite:`Roberts2004`.
+
+    Attributes
+    ----------
+    kmer_alphabet : KmerAlphabet
+        The *k-mer* alphabet.
+    window : int
+        The window size.
+    permutation : Permutation
+        The permutation.
+
+    Notes
+    -----
+    For minimizer computation a fast algorithm :footcite:`VanHerk1992`
+    is used, whose runtime scales linearly with the length of the
+    sequence and is constant with regard to the size of the rolling
+    window.
+
+    References
+    ----------
+
+    .. footbibliography::
+
+    Examples
+    --------
+
+    The *k-mer* decomposition of a sequence can yield a high number of
+    *k-mers*:
+
+    >>> sequence1 = ProteinSequence("THIS*IS*A*SEQVENCE")
+    >>> kmer_alph = KmerAlphabet(sequence1.alphabet, k=3)
+    >>> all_kmers = kmer_alph.create_kmers(sequence1.code)
+    >>> print(all_kmers)
+    [ 9367  3639  4415  9199 13431  4415  9192 13271   567 13611  8725  2057
+      7899  9875  1993  6363]
+    >>> print(["".join(kmer_alph.decode(kmer)) for kmer in all_kmers])
+    ['THI', 'HIS', 'IS*', 'S*I', '*IS', 'IS*', 'S*A', '*A*', 'A*S', '*SE', 'SEQ', 'EQV', 'QVE', 'VEN', 'ENC', 'NCE']
+
+    Minimizers can be used to reduce the number of *k-mers* by selecting
+    only the minimum *k-mer* in each window *w*:
+
+    >>> minimizer = MinimizerSelector(kmer_alph, window=4)
+    >>> minimizer_pos, minimizers = minimizer.select(sequence1)
+    >>> print(minimizer_pos)
+    [ 1  2  5  8 11 14]
+    >>> print(minimizers)
+    [3639 4415 4415  567 2057 1993]
+    >>> print(["".join(kmer_alph.decode(kmer)) for kmer in minimizers])
+    ['HIS', 'IS*', 'IS*', 'A*S', 'EQV', 'ENC']
+
+    Although this approach reduces the number of *k-mers*, minimizers
+    are still guaranteed to match minimizers in another sequence, if
+    they share an equal subsequence of at least length *w + k - 1*:
+
+    >>> sequence2 = ProteinSequence("ANQTHER*SEQVENCE")
+    >>> other_minimizer_pos, other_minimizers = minimizer.select(sequence2)
+    >>> print(["".join(kmer_alph.decode(kmer)) for kmer in other_minimizers])
+    ['ANQ', 'HER', 'ER*', 'EQV', 'ENC']
+    >>> common_minimizers = set.intersection(set(minimizers), set(other_minimizers))
+    >>> print(["".join(kmer_alph.decode(kmer)) for kmer in common_minimizers])
+    ['EQV', 'ENC']
+    """
+
+    def __init__(self, kmer_alphabet, window, permutation=None):
+        if window < 2:
+            raise ValueError("Window size must be at least 2")
+        self._window = window
+        self._kmer_alph = kmer_alphabet
+        self._permutation = permutation
+
+
+    @property
+    def kmer_alphabet(self):
+        return self._kmer_alph
+
+    @property
+    def window(self):
+        return self._window
+
+    @property
+    def permutation(self):
+        return self._permutation
+
+
+    def select(self, sequence, bint alphabet_check=True):
+        """
+        select(sequence, alphabet_check=True)
+
+        Obtain all overlapping *k-mers* from a sequence and select
+        the minimizers from them.
+
+        Parameters
+        ----------
+        sequence : Sequence
+            The sequence to find the minimizers in.
+            Must be compatible with the given `kmer_alphabet`
+        alphabet_check: bool, optional
+            If set to false, the compatibility between the alphabet
+            of the sequence and the alphabet of the
+            :class:`MinimizerSelector`
+            is not checked to gain additional performance.
+
+        Returns
+        -------
+        minimizer_indices : ndarray, dtype=np.uint32
+            The sequence indices where the minimizer *k-mers* start.
+        minimizers : ndarray, dtype=np.int64
+            The *k-mers* that are the selected minimizers, returned as
+            *k-mer* code.
+
+        Notes
+        -----
+        Duplicate minimizers are omitted, i.e. if two windows have the
+        same minimizer position, the return values contain this
+        minimizer only once.
+        """
+        if alphabet_check:
+            if not self._kmer_alph.base_alphabet.extends(sequence.alphabet):
+                raise ValueError(
+                    "The sequence's alphabet does not fit the k-mer alphabet"
+                )
+        kmers = self._kmer_alph.create_kmers(sequence.code)
+        return self.select_from_kmers(kmers)
+
+
+    def select_from_kmers(self, kmers):
+        """
+        select_from_kmers(kmers)
+
+        Select minimizers for the given overlapping *k-mers*.
+
+        Parameters
+        ----------
+        kmers : ndarray, dtype=np.int64
+            The *k-mer* codes representing the sequence to find the
+            minimizers in.
+            The *k-mer* codes correspond to the *k-mers* encoded by the
+            given `kmer_alphabet`.
+
+        Returns
+        -------
+        minimizer_indices : ndarray, dtype=np.uint32
+            The indices in the input *k-mer* sequence where a minimizer
+            appears.
+        minimizers : ndarray, dtype=np.int64
+            The corresponding *k-mers* codes of the minimizers.
+
+        Notes
+        -----
+        Duplicate minimizers are omitted, i.e. if two windows have the
+        same minimizer position, the return values contain this
+        minimizer only once.
+        """
+        if self._permutation is None:
+            ordering = kmers
+        else:
+            ordering = self._permutation.permute(kmers)
+            if len(ordering) != len(kmers):
+                raise IndexError(
+                    f"The Permutation is defective, it gave {len(ordering)} "
+                    f"sort keys for {len(kmers)} k-mers"
+                )
+
+        if len(kmers) < self._window:
+            raise ValueError(
+                "The number of k-mers is smaller than the window size"
+            )
+        return _minimize(
+            kmers.astype(np.int64, copy=False),
+            ordering.astype(np.int64, copy=False),
+            self._window,
+            include_duplicates=False
+        )
+
+
+class SyncmerSelector:
+    """
+    SyncmerSelector(alphabet, k, s, permutation=None, offset=(0,))
+
+    Selects the *syncmers* in sequences.
+
+    Let the *s-mers* be all overlapping substrings of length *s* in a
+    *k-mer*.
+    A *k-mer* is a syncmer, if its minimum *s-mer* is at one of the
+    given offset positions :footcite:`Edgar2021`.
+    If the same minimum *s-mer* appears twice in a *k-mer*, the position
+    of the leftmost *s-mer* is taken.
+
+    Parameters
+    ----------
+    alphabet : Alphabet
+        The base alphabet the *k-mers* and *s-mers* are created from.
+        Defines the type of sequence this :class:`MinimizerSelector` can
+        be applied on.
+    k, s : int
+        The length of the *k-mers* and *s-mers*, respectively.
+    permutation : Permutation
+        If set, the *s-mer* order is permuted, i.e.
+        the minimum *s-mer* is chosen based on the ordering of the sort
+        keys from :class:`Permutation.permute()`.
+        This :class:`Permutation` must be compatible with *s*
+        (not with *k*).
+        By default, the standard order of the :class:`KmerAlphabet` is
+        used.
+        This standard order is often the lexicographical order, which is
+        known to yield suboptimal *density* in many cases
+        :footcite:`Roberts2004`.
+    offset : array-like of int
+        If the minimum *s-mer* in a *k-mer* is at one of the given
+        offset positions, that *k-mer* is a syncmer.
+        Negative values indicate the position from the end of the
+        *k-mer*.
+        By default, the minimum position needs to be at the start of the
+        *k-mer*, which is termed *open syncmer*.
+
+    Attributes
+    ----------
+    alphabet : Alphabet
+        The base alphabet.
+    kmer_alphabet, smer_alphabet : int
+        The :class:`KmerAlphabet` for *k* and *s*, respectively.
+    permutation : Permutation
+        The permutation.
+
+    See Also
+    --------
+    CachedSyncmerSelector
+        A cached variant with faster syncmer selection at the cost of
+        increased initialization time.
+
+    Notes
+    -----
+    For syncmer computation from a sequence a fast algorithm
+    :footcite:`VanHerk1992` is used, whose runtime scales linearly with
+    the length of the sequence and is constant with regard to *k*.
+
+    References
+    ----------
+
+    .. footbibliography::
+
+    Examples
+    --------
+
+    This example is taken from :footcite:`Edgar2021`:
+    The subset of *k-mers* that are *closed syncmers* are selected.
+    Closed syncmers are syncmers, where the minimum *s-mer* is in the
+    first or last position of the *k-mer*.
+    *s-mers* are ordered lexicographically in this example.
+
+    >>> sequence = NucleotideSequence("GGCAAGTGACA")
+    >>> kmer_alph = KmerAlphabet(sequence.alphabet, k=5)
+    >>> kmers = kmer_alph.create_kmers(sequence.code)
+    >>> closed_syncmer_selector = CachedSyncmerSelector(
+    ...     sequence.alphabet,
+    ...     # The same k as in the KmerAlphabet
+    ...     k=5,
+    ...     s=2,
+    ...     # The offset determines that closed syncmers will be selected
+    ...     offset=(0, -1)
+    ... )
+    >>> syncmer_pos, syncmers = closed_syncmer_selector.select(sequence)
+    >>> # Print all k-mers in the sequence and mark syncmers with a '*'
+    >>> for pos, kmer in enumerate(kmer_alph.create_kmers(sequence.code)):
+    ...     if pos in syncmer_pos:
+    ...         print("* " + "".join(kmer_alph.decode(kmer)))
+    ...     else:
+    ...         print("  " + "".join(kmer_alph.decode(kmer)))
+    * GGCAA
+      GCAAG
+      CAAGT
+    * AAGTG
+    * AGTGA
+    * GTGAC
+      TGACA
+    """
+
+    def __init__(self, alphabet, k, s, permutation=None, offset=(0,)):
+        if not s < k:
+            raise ValueError("s must be smaller than k")
+        self._window = k - s + 1
+        self._alphabet = alphabet
+        self._kmer_alph = KmerAlphabet(alphabet, k)
+        self._smer_alph = KmerAlphabet(alphabet, s)
+
+        self._permutation = permutation
+
+        self._offset = np.asarray(offset, dtype=np.int64)
+        # Wrap around negative indices
+        self._offset = np.where(
+            self._offset < 0,
+            self._window + self._offset,
+            self._offset
+        )
+        if (self._offset >= self._window).any() or (self._offset < 0).any():
+            raise IndexError(
+                f"Offset is out of window range"
+            )
+        if len(np.unique(self._offset)) != len(self._offset):
+            raise ValueError("Offset must contain unique values")
+
+
+    @property
+    def alphabet(self):
+        return self._alphabet
+
+    @property
+    def kmer_alphabet(self):
+        return self._kmer_alph
+
+    @property
+    def smer_alphabet(self):
+        return self._smer_alph
+
+    @property
+    def permutation(self):
+        return self._permutation
+
+
+    def select(self, sequence, bint alphabet_check=True):
+        """
+        select(sequence, alphabet_check=True)
+
+        Obtain all overlapping *k-mers* from a sequence and select
+        the syncmers from them.
+
+        Parameters
+        ----------
+        sequence : Sequence
+            The sequence to find the syncmers in.
+            Must be compatible with the given `kmer_alphabet`
+        alphabet_check: bool, optional
+            If set to false, the compatibility between the alphabet
+            of the sequence and the alphabet of the
+            :class:`SyncmerSelector`
+            is not checked to gain additional performance.
+
+        Returns
+        -------
+        syncmer_indices : ndarray, dtype=np.uint32
+            The sequence indices where the syncmers start.
+        syncmers : ndarray, dtype=np.int64
+            The corresponding *k-mer* codes of the syncmers.
+        """
+        if alphabet_check:
+            if not self._alphabet.extends(sequence.alphabet):
+                raise ValueError(
+                    "The sequence's alphabet does not fit "
+                    "the selector's alphabet"
+                )
+        kmers = self._kmer_alph.create_kmers(sequence.code)
+        smers = self._smer_alph.create_kmers(sequence.code)
+
+        if self._permutation is None:
+            ordering = smers
+        else:
+            ordering = self._permutation.permute(smers)
+            if len(ordering) != len(smers):
+                raise IndexError(
+                    f"The Permutation is defective, it gave {len(ordering)} "
+                    f"sort keys for {len(smers)} s-mers"
+                )
+
+        # The aboslute position of the minimum s-mer for each k-mer
+        min_pos, _ = _minimize(
+            smers,
+            ordering.astype(np.int64, copy=False),
+            self._window,
+            include_duplicates=True
+        )
+        # The position of the minimum s-mer relative to the start
+        # of the k-mer
+        relative_min_pos = min_pos - np.arange(len(kmers))
+        syncmer_pos = self._filter_syncmer_pos(relative_min_pos)
+        return syncmer_pos, kmers[syncmer_pos]
+
+
+    def select_from_kmers(self, kmers):
+        """
+        select_from_kmers(kmers)
+
+        Select syncmers for the given *k-mers*.
+
+        The *k-mers* are not required to overlap.
+
+        Parameters
+        ----------
+        kmers : ndarray, dtype=np.int64
+            The *k-mer* codes to select the syncmers from.
+
+        Returns
+        -------
+        syncmer_indices : ndarray, dtype=np.uint32
+            The sequence indices where the syncmers start.
+        syncmers : ndarray, dtype=np.int64
+            The corresponding *k-mer* codes of the syncmers.
+
+        Notes
+        -----
+        Since for *s-mer* creation, the *k-mers* need to be converted
+        back to symbol codes again and since the input *k-mers* are not
+        required to overlap, calling :meth:`select()` is much faster.
+        However, :meth:`select()` is only available for
+        :class:`Sequence` objects.
+        """
+        cdef int64 i
+
+        symbol_codes_for_each_kmer = self._kmer_alph.split(kmers)
+
+        cdef int64[:] min_pos = np.zeros(
+            len(symbol_codes_for_each_kmer), dtype=np.int64
+        )
+        for i in range(symbol_codes_for_each_kmer.shape[0]):
+            smers = self._smer_alph.create_kmers(symbol_codes_for_each_kmer[i])
+            if self._permutation is None:
+                ordering = smers
+            else:
+                ordering = self._permutation.permute(smers)
+                if len(ordering) != len(smers):
+                    raise IndexError(
+                        f"The Permutation is defective, it gave {len(ordering)} "
+                        f"sort keys for {len(smers)} s-mers"
+                    )
+            min_pos[i] = np.argmin(ordering)
+
+        syncmer_pos = self._filter_syncmer_pos(min_pos)
+        return syncmer_pos, kmers[syncmer_pos]
+
+
+    def _filter_syncmer_pos(self, min_pos):
+        """
+        Get indices of *k-mers* that are syncmers, based on `min_pos`,
+        the position of the minimum *s-mer* in each *k-mer*.
+        Syncmers are k-mers whose the minimum s-mer is at (one of)
+        the given offet position(s).
+        """
+        syncmer_mask = None
+        for offset in self._offset:
+            # For the usual number of offsets, this 'loop'-appoach is
+            # faster than np.isin()
+            if syncmer_mask is None:
+                syncmer_mask = min_pos == offset
+            else:
+                syncmer_mask |= min_pos == offset
+        return np.where(syncmer_mask)[0]
+
+
+class CachedSyncmerSelector(SyncmerSelector):
+    """
+    CachedSyncmerSelector(alphabet, k, s, permutation=None, offset=(0,))
+
+    Selects the *syncmers* in sequences.
+
+    Fulsfills the same purpose as :class:`SyncmerSelector`, but
+    precomputes for each possible *k-mer*, whether it is a syncmer,
+    at initialization.
+    Hence, syncmer selection is faster at the cost of longer
+    initialization time.
+
+    Parameters
+    ----------
+    alphabet : Alphabet
+        The base alphabet the *k-mers* and *s-mers* are created from.
+        Defines the type of sequence this :class:`MinimizerSelector` can
+        be applied on.
+    k, s : int
+        The length of the *k-mers* and *s-mers*, respectively.
+    permutation : Permutation
+        If set, the *s-mer* order is permuted, i.e.
+        the minimum *s-mer* is chosen based on the ordering of the sort
+        keys from :class:`Permutation.permute()`.
+        This :class:`Permutation` must be compatible with *s*
+        (not with *k*).
+        By default, the standard order of the :class:`KmerAlphabet` is
+        used.
+        This standard order is often the lexicographical order, which is
+        known to yield suboptimal *density* in many cases
+        :footcite:`Roberts2004`.
+    offset : array-like of int
+        If the minimum *s-mer* in a *k-mer* is at one of the given
+        offset positions, that *k-mer* is a syncmer.
+        Negative values indicate the position from the end of the
+        *k-mer*.
+        By default, the minimum position needs to be at the start of the
+        *k-mer*, which is termed *open syncmer*.
+
+    Attributes
+    ----------
+    alphabet : Alphabet
+        The base alphabet.
+    kmer_alphabet, smer_alphabet : int
+        The :class:`KmerAlphabet` for *k* and *s*, respectively.
+    permutation : Permutation
+        The permutation.
+
+    See Also
+    --------
+    SyncmerSelector
+        A standard variant for syncmer selection.
+
+    Notes
+    -----
+    Both the initialization time and memory requirements are
+    proportional to the size of the `kmer_alphabet`, i.e. :math:`n^k`.
+    Hence, it is adviced to use this class only for rather small
+    alphabets.
+
+    References
+    ----------
+
+    .. footbibliography::
+
+    Examples
+    --------
+
+    >>> sequence = NucleotideSequence("GGCAAGTGACA")
+    >>> kmer_alph = KmerAlphabet(sequence.alphabet, k=5)
+    >>> # The initialization can quite a long time for large *k-mer* alphabets...
+    >>> closed_syncmer_selector = CachedSyncmerSelector(
+    ...     sequence.alphabet,
+    ...     # The same k as in the KmerAlphabet
+    ...     k=5,
+    ...     s=2,
+    ...     # The offset determines that closed syncmers will be selected
+    ...     offset=(0, -1)
+    ... )
+    >>> # ...but the actual syncmer identification is very fast
+    >>> syncmer_pos, syncmers = closed_syncmer_selector.select(sequence)
+    >>> print(["".join(kmer_alph.decode(kmer)) for kmer in syncmers])
+    ['GGCAA', 'AAGTG', 'AGTGA', 'GTGAC']
+    """
+
+    def __init__(self, alphabet, k, s, permutation=None, offset=(0,)):
+        super().__init__(alphabet, k, s, permutation, offset)
+        # Check for all possible *k-mers*, whether they are syncmers
+        all_kmers = np.arange(len(self.kmer_alphabet))
+        syncmer_indices, _ = super().select_from_kmers(all_kmers)
+        # Convert the index array into a boolean mask
+        self._syncmer_mask = np.zeros(len(self.kmer_alphabet), dtype=bool)
+        self._syncmer_mask[syncmer_indices] = True
+
+
+    def select(self, sequence, bint alphabet_check=True):
+        """
+        select(sequence, alphabet_check=True)
+
+        Obtain all overlapping *k-mers* from a sequence and select
+        the syncmers from them.
+
+        Parameters
+        ----------
+        sequence : Sequence
+            The sequence to find the syncmers in.
+            Must be compatible with the given `kmer_alphabet`
+        alphabet_check: bool, optional
+            If set to false, the compatibility between the alphabet
+            of the sequence and the alphabet of the
+            :class:`CachedSyncmerSelector`
+            is not checked to gain additional performance.
+
+        Returns
+        -------
+        syncmer_indices : ndarray, dtype=np.uint32
+            The sequence indices where the syncmers start.
+        syncmers : ndarray, dtype=np.int64
+            The corresponding *k-mer* codes of the syncmers.
+        """
+        if alphabet_check:
+            if not self.alphabet.extends(sequence.alphabet):
+                raise ValueError(
+                    "The sequence's alphabet does not fit "
+                    "the selector's alphabet"
+                )
+        kmers = self.kmer_alphabet.create_kmers(sequence.code)
+        return self.select_from_kmers(kmers)
+
+
+    def select_from_kmers(self, kmers):
+        """
+        select_from_kmers(kmers)
+
+        Select syncmers for the given *k-mers*.
+
+        The *k-mers* are not required to overlap.
+
+        Parameters
+        ----------
+        kmers : ndarray, dtype=np.int64
+            The *k-mer* codes to select the syncmers from.
+
+        Returns
+        -------
+        syncmer_indices : ndarray, dtype=np.uint32
+            The sequence indices where the syncmers start.
+        syncmers : ndarray, dtype=np.int64
+            The corresponding *k-mer* codes of the syncmers.
+        """
+        syncmer_pos = np.where(self._syncmer_mask[kmers])[0]
+        return syncmer_pos, kmers[syncmer_pos]
+
+
+class MincodeSelector:
+    r"""
+    MincodeSelector(self, kmer_alphabet, compression, permutation=None)
+
+    Selects the :math:`1/\text{compression}` *smallest* *k-mers* from
+    :class:`KmerAlphabet`. :footcite:`Edgar2021`
+
+    '*Small*' refers to the lexicographical order, or alternatively a
+    custom order if `permutation` is given.
+    The *Mincode* approach tries to reduce the number of *k-mers* from a
+    sequence by the factor `compression`, while it still ensures that
+    a common set of *k-mers* are selected from similar sequences.
+
+    Parameters
+    ----------
+    kmer_alphabet : KmerAlphabet
+        The *k-mer* alphabet that defines the *k-mer* size and the type
+        of sequence this :class:`MincodeSelector` can be applied on.
+    compression : float
+        Defines the compression factor, i.e. the approximate fraction
+        of *k-mers* that will be sampled from a sequence.
+    permutation : Permutation
+        If set, the *k-mer* order is permuted, i.e.
+        the *k-mers* are selected based on the ordering of the sort keys
+        from :class:`Permutation.permute()`.
+        By default, the standard order of the :class:`KmerAlphabet` is
+        used.
+        This standard order is often the lexicographical order.
+
+    Attributes
+    ----------
+    kmer_alphabet : KmerAlphabet
+        The *k-mer* alphabet.
+    compression : float
+        The compression factor.
+    threshold : float
+        Based on the compression factor and the range of (permuted)
+        *k-mer* values this threshold is calculated.
+        All *k-mers*, that are smaller than this value are selected.
+    permutation : Permutation
+        The permutation.
+
+    References
+    ----------
+
+    .. footbibliography::
+
+    Examples
+    --------
+
+    >>> kmer_alph = KmerAlphabet(NucleotideSequence.alphabet_unamb, k=2)
+    >>> kmers = np.arange(len(kmer_alph))
+    >>> print(["".join(kmer_alph.decode(kmer)) for kmer in kmers])
+    ['AA', 'AC', 'AG', 'AT', 'CA', 'CC', 'CG', 'CT', 'GA', 'GC', 'GG', 'GT', 'TA', 'TC', 'TG', 'TT']
+    >>> # Select 1/4 of *k-mers* based on lexicographical k-mer order
+    >>> selector = MincodeSelector(kmer_alph, 4)
+    >>> subset_pos, kmers_subset = selector.select_from_kmers(kmers)
+    >>> print(["".join(kmer_alph.decode(kmer)) for kmer in kmers_subset])
+    ['AA', 'AC', 'AG', 'AT']
+    >>> # Select 1/4 based on randomized k-mer order
+    >>> selector = MincodeSelector(kmer_alph, 4, permutation=RandomPermutation())
+    >>> subset_pos, kmers_subset = selector.select_from_kmers(kmers)
+    >>> print(["".join(kmer_alph.decode(kmer)) for kmer in kmers_subset])
+    ['AG', 'CT', 'GA', 'TC']
+    """
+
+    def __init__(self, kmer_alphabet, compression, permutation=None):
+        if compression < 1:
+            raise ValueError(
+                "Compression factor must be equal to or larger than 1"
+            )
+        self._compression = compression
+        self._kmer_alph = kmer_alphabet
+        self._permutation = permutation
+        if permutation is None:
+            permutation_offset = 0
+            permutation_range = len(kmer_alphabet)
+        else:
+            permutation_offset = permutation.min
+            permutation_range = permutation.max - permutation.min + 1
+        self._threshold = permutation_offset + permutation_range / compression
+
+
+    @property
+    def kmer_alphabet(self):
+        return self._kmer_alph
+
+    @property
+    def compression(self):
+        return self._compression
+
+    @property
+    def threshold(self):
+        return self._threshold
+
+    @property
+    def permutation(self):
+        return self._permutation
+
+
+    def select(self, sequence, bint alphabet_check=True):
+        """
+        select(sequence, alphabet_check=True)
+
+        Obtain all overlapping *k-mers* from a sequence and select
+        the *Mincode k-mers* from them.
+
+        Parameters
+        ----------
+        sequence : Sequence
+            The sequence to find the *Mincode k-mers* in.
+            Must be compatible with the given `kmer_alphabet`
+        alphabet_check: bool, optional
+            If set to false, the compatibility between the alphabet
+            of the sequence and the alphabet of the
+            :class:`MincodeSelector`
+            is not checked to gain additional performance.
+
+        Returns
+        -------
+        mincode_indices : ndarray, dtype=np.uint32
+            The sequence indices where the *Mincode k-mers* start.
+        mincode : ndarray, dtype=np.int64
+            The corresponding *Mincode k-mer* codes.
+        """
+        if alphabet_check:
+            if not self._kmer_alph.base_alphabet.extends(sequence.alphabet):
+                raise ValueError(
+                    "The sequence's alphabet does not fit the k-mer alphabet"
+                )
+        kmers = self._kmer_alph.create_kmers(sequence.code)
+        return self.select_from_kmers(kmers)
+
+
+    def select_from_kmers(self, kmers):
+        """
+        select_from_kmers(kmers)
+
+        Select *Mincode k-mers*.
+
+        The given *k-mers* are not required to overlap.
+
+        Parameters
+        ----------
+        kmers : ndarray, dtype=np.int64
+            The *k-mer* codes to select the *Mincode k-mers* from.
+
+        Returns
+        -------
+        mincode_indices : ndarray, dtype=np.uint32
+            The sequence indices where the *Mincode k-mers* start.
+        mincode : ndarray, dtype=np.int64
+            The corresponding *Mincode k-mer* codes.
+        """
+        if self._permutation is None:
+            ordering = kmers
+        else:
+            ordering = self._permutation.permute(kmers)
+            if len(ordering) != len(kmers):
+                raise IndexError(
+                    f"The Permutation is defective, it gave {len(ordering)} "
+                    f"sort keys for {len(kmers)} k-mers"
+                )
+
+        mincode_pos = ordering < self._threshold
+        return mincode_pos, kmers[mincode_pos]
+
+
+@cython.boundscheck(False)
+@cython.wraparound(False)
+def _minimize(int64[:] kmers, int64[:] ordering, uint32 window,
+              bint include_duplicates):
+    """
+    Implementation of the algorithm originally devised by
+    Marcel van Herk.
+
+    In this implementation the frame is chosen differently:
+    For a position 'x' the frame ranges from 'x' to 'x + window-1'
+    instead of 'x - (window-1)/2' to 'x + (window-1)/2'.
+    """
+    cdef uint32 seq_i
+
+    cdef uint32 n_windows = kmers.shape[0] - (window - 1)
+    # Pessimistic array allocation size
+    # -> Expect that every window has a new minimizer
+    cdef uint32[:] mininizer_pos = np.empty(n_windows, dtype=np.uint32)
+    cdef int64[:] minimizers = np.empty(n_windows, dtype=np.int64)
+    # Counts the actual number of minimiers for later trimming
+    cdef uint32 n_minimizers = 0
+
+    # Variables for the position of the previous cumulative minimum
+    # Assign an value that can never occur for the start,
+    # as in the beginning there is no previous value
+    cdef uint32 prev_argcummin = kmers.shape[0]
+    # Variables for the position of the current cumulative minimum
+    cdef uint32 combined_argcummin, forward_argcummin, reverse_argcummin
+    # Variables for the current cumulative minimum
+    cdef int64 combined_cummin, forward_cummin, reverse_cummin
+    # Variables for cumulative minima at all positions
+    cdef uint32[:] forward_argcummins = _chunk_wise_forward_argcummin(
+        ordering, window
+    )
+    cdef uint32[:] reverse_argcummins = _chunk_wise_reverse_argcummin(
+        ordering, window
+    )
+
+    for seq_i in range(n_windows):
+        forward_argcummin = forward_argcummins[seq_i + window - 1]
+        reverse_argcummin = reverse_argcummins[seq_i]
+        forward_cummin = ordering[forward_argcummin]
+        reverse_cummin = ordering[reverse_argcummin]
+
+        # At ties the leftmost position is taken,
+        # which stems from the reverse pass
+        if forward_cummin < reverse_cummin:
+            combined_argcummin = forward_argcummin
+        else:
+            combined_argcummin = reverse_argcummin
+
+        # If the same minimizer position was observed before, the
+        # duplicate is simply ignored, if 'include_duplicates' is false
+        if include_duplicates or combined_argcummin != prev_argcummin:
+            # Append minimizer to return value
+            mininizer_pos[n_minimizers] = combined_argcummin
+            minimizers[n_minimizers] = kmers[combined_argcummin]
+            n_minimizers += 1
+            prev_argcummin = combined_argcummin
+
+    return (
+        np.asarray(mininizer_pos)[:n_minimizers],
+        np.asarray(minimizers)[:n_minimizers]
+    )
+
+@cython.boundscheck(False)
+@cython.wraparound(False)
+@cython.cdivision(True)
+cdef _chunk_wise_forward_argcummin(int64[:] values, uint32 chunk_size):
+    """
+    Argument of the cumulative minimum.
+    """
+    cdef uint32 seq_i
+
+    cdef uint32 current_min_i = 0
+    cdef int64 current_min, current_val
+    cdef uint32[:] min_pos = np.empty(values.shape[0], dtype=np.uint32)
+
+    # Any actual value will be smaller than this placeholder
+    current_min = MAX_INT_64
+    for seq_i in range(values.shape[0]):
+        if seq_i % chunk_size == 0:
+            # New chunk begins
+            current_min = MAX_INT_64
+        current_val = values[seq_i]
+        if current_val < current_min:
+            current_min_i = seq_i
+            current_min = current_val
+        min_pos[seq_i] = current_min_i
+
+    return min_pos
+
+@cython.boundscheck(False)
+@cython.wraparound(False)
+@cython.cdivision(True)
+cdef _chunk_wise_reverse_argcummin(int64[:] values, uint32 chunk_size):
+    """
+    The same as above but starting from the other end and iterating
+    backwards.
+    Separation into two functions leads to code duplication.
+    However, single implemention with reversed `values` as input
+    has some disadvantages:
+
+    - Indices must be transformed so that they point to the
+      non-reversed `values`
+    - There are issues in selecting the leftmost argument
+    - An offset is necessary to ensure alignment of chunks with forward
+      pass
+
+    Hence, a separate 'reverse' variant of the function was implemented.
+    """
+    cdef uint32 seq_i
+
+    cdef uint32 current_min_i = 0
+    cdef int64 current_min, current_val
+    cdef uint32[:] min_pos = np.empty(values.shape[0], dtype=np.uint32)
+
+    current_min = MAX_INT_64
+    for seq_i in reversed(range(values.shape[0])):
+        # The chunk beginning is a small difference to forward
+        # implementation, as it begins on the left of the chunk border
+        if seq_i % chunk_size == chunk_size - 1:
+            current_min = MAX_INT_64
+        current_val = values[seq_i]
+        # The '<=' is a small difference to forward implementation
+        # to enure the loftmost argument is selected
+        if current_val <= current_min:
+            current_min_i = seq_i
+            current_min = current_val
+        min_pos[seq_i] = current_min_i
+
+    return min_pos