Merge pull request #373 from padix-key/msa

Add support for Muscle 5

.github/workflows/test.yml

@@ -85,3 +85,32 @@ jobs:
         run: pip install .//dist//*.whl
       - name: Testing code
         run: pytest --ignore=tests//application//test_blast.py --ignore=tests//application//test_sra.py --ignore=tests//database//test_entrez.py
+
+
+  test-muscle5:
+    name: Testing Biotite's interface for Muscle 5
+
+    runs-on: ubuntu-latest
+    defaults:
+      run:
+        shell: bash -l {0}
+
+    steps:
+      - uses: actions/checkout@v2
+      - uses: conda-incubator/setup-miniconda@v2
+        with:
+          activate-environment: biotite-dev
+          auto-update-conda: true
+          python-version: 3.9
+          mamba-version: "*"
+          channels: conda-forge,defaults
+      - name: Installing dependencies
+        run: conda install -c conda-forge -c bioconda "muscle=5" "cython>=0.29" "numpy=1.19" "requests>=2.12" "msgpack-python>=0.5.6" "networkx>=2.0" "pytest>=3.2"
+        env:
+          NUMPY_VERSION: ${{ matrix.numpy }}
+      - name: Building distribution
+        run: python setup.py bdist_wheel
+      - name: Installing distribution
+        run: pip install .//dist//*.whl
+      - name: Testing code
+        run: pytest tests//application//test_msa.py

environment.yml

@@ -32,7 +32,7 @@ dependencies:
   - dssp
   - clustalo
   - mafft
-  - muscle
+  - muscle =3
   - sra-tools
   - autodock-vina
   - viennarna

src/biotite/application/application.py

@@ -4,8 +4,8 @@
 
 __name__ = "biotite.application"
 __author__ = "Patrick Kunzmann"
-__all__ = ["Application", "AppStateError", "TimeoutError", "AppState",
-           "requires_state"]
+__all__ = ["Application", "AppStateError", "TimeoutError", "VersionError",
+           "AppState", "requires_state"]
 
 import abc
 import time
@@ -251,3 +251,10 @@ class TimeoutError(Exception):
     Indicate that the application's timeout expired.
     """
     pass
+
+
+class VersionError(Exception):
+    """
+    Indicate that the application's version is invalid.
+    """
+    pass

src/biotite/application/muscle/__init__.py

@@ -9,4 +9,5 @@
 __name__ = "biotite.application.muscle"
 __author__ = "Patrick Kunzmann"
 
-from .app import *
+from .app3 import *
+from .app5 import *

src/biotite/application/muscle/app.py → src/biotite/application/muscle/app3.py

@@ -6,12 +6,14 @@
 __author__ = "Patrick Kunzmann"
 __all__ = ["MuscleApp"]
 
+import re
 import numbers
 import warnings
+import subprocess
 from tempfile import NamedTemporaryFile
 from ..localapp import cleanup_tempfile
 from ..msaapp import MSAApp
-from ..application import AppState, requires_state
+from ..application import AppState, VersionError, requires_state
 from ...sequence.sequence import Sequence
 from ...sequence.seqtypes import NucleotideSequence, ProteinSequence
 from ...sequence.align.matrix import SubstitutionMatrix
@@ -21,7 +23,7 @@
 
 class MuscleApp(MSAApp):
     """
-    Perform a multiple sequence alignment using MUSCLE.
+    Perform a multiple sequence alignment using MUSCLE version 3.
     
     Parameters
     ----------
@@ -32,6 +34,10 @@ class MuscleApp(MSAApp):
     matrix : SubstitutionMatrix, optional
         A custom substitution matrix.
     
+    See also
+    ---------
+    Muscle5App
+    
     Examples
     --------
 
@@ -51,6 +57,12 @@ class MuscleApp(MSAApp):
     """
 
     def __init__(self, sequences, bin_path="muscle", matrix=None):
+        major_version = get_version(bin_path)[0]
+        if major_version != 3:
+            raise VersionError(
+                f"Muscle 3 is required, got version {major_version}"
+            )
+
         super().__init__(sequences, bin_path, matrix)
         self._gap_open = None
         self._gap_ext = None
@@ -67,8 +79,8 @@ def __init__(self, sequences, bin_path="muscle", matrix=None):
     def run(self):
         args = [
             "-quiet",
-            "-in",    self.get_input_file_path(),
-            "-out",   self.get_output_file_path(),
+            "-in", self.get_input_file_path(),
+            "-out", self.get_output_file_path(),
             "-tree1", self._out_tree1_file.name,
             "-tree2", self._out_tree2_file.name,
         ]
@@ -191,7 +203,7 @@ def align(cls, sequences, bin_path=None, matrix=None,
         """
         Perform a multiple sequence alignment.
         
-        This is a convenience function, that wraps the :class:`MSAApp`
+        This is a convenience function, that wraps the :class:`MuscleApp`
         execution.
         
         Parameters
@@ -225,3 +237,16 @@ def align(cls, sequences, bin_path=None, matrix=None,
         app.start()
         app.join()
         return app.get_alignment()
+
+
+def get_version(bin_path="muscle"):
+    output = subprocess.run(
+        [bin_path, "-version"], capture_output=True, text=True
+    )
+    # Find matches for version string containing major and minor version
+    match = re.search("\d+\.\d+", output.stdout)
+    if match is None:
+        raise subprocess.SubprocessError("Application did not print version")
+    version_string = match.group(0)
+    splitted = version_string.split(".")
+    return int(splitted[0]), int(splitted[1])

src/biotite/application/muscle/app5.py

@@ -0,0 +1,171 @@
+# 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.application.muscle"
+__author__ = "Patrick Kunzmann"
+__all__ = ["Muscle5App"]
+
+import numbers
+import warnings
+from tempfile import NamedTemporaryFile
+from ..localapp import cleanup_tempfile
+from ..msaapp import MSAApp
+from ..application import AppState, VersionError, requires_state
+from ...sequence.sequence import Sequence
+from ...sequence.seqtypes import NucleotideSequence, ProteinSequence
+from ...sequence.align.matrix import SubstitutionMatrix
+from ...sequence.align.alignment import Alignment
+from ...sequence.phylo.tree import Tree
+from .app3 import get_version
+
+
+class Muscle5App(MSAApp):
+    """
+    Perform a multiple sequence alignment using MUSCLE version 5.
+    
+    Parameters
+    ----------
+    sequences : list of Sequence
+        The sequences to be aligned.
+    bin_path : str, optional
+        Path of the MUSCLE binary.
+    
+    See also
+    ---------
+    MuscleApp
+
+    Notes
+    -----
+    Alignment ensemble generation is not supported, yet.
+    
+    Examples
+    --------
+
+    >>> seq1 = ProteinSequence("BIQTITE")
+    >>> seq2 = ProteinSequence("TITANITE")
+    >>> seq3 = ProteinSequence("BISMITE")
+    >>> seq4 = ProteinSequence("IQLITE")
+    >>> app = Muscle5App([seq1, seq2, seq3, seq4])
+    >>> app.start()
+    >>> app.join()
+    >>> alignment = app.get_alignment()
+    >>> print(alignment)
+    BI-QTITE
+    TITANITE
+    BI-SMITE
+    -I-QLITE
+    """
+
+    def __init__(self, sequences, bin_path="muscle"):
+        major_version = get_version(bin_path)[0]
+        if major_version < 5:
+            raise VersionError(
+                f"At least Muscle 5 is required, got version {major_version}"
+            )
+
+        super().__init__(sequences, bin_path)
+        self._mode = "align"
+        self._consiters = None
+        self._refineiters = None
+        self._n_threads = None
+
+    @requires_state(AppState.CREATED)
+    def set_iterations(self, consistency=None, refinement=None):
+        """
+        Set the number of iterations for the alignment algorithm.
+
+        Parameters
+        ----------
+        consistency : int, optional
+            The number of consistency iterations.
+        refinement : int, optional
+            The number of refinement iterations.
+        """
+        if consistency is not None:
+            self._consiters = consistency
+        if refinement is not None:
+            self._refineiters = refinement
+
+    @requires_state(AppState.CREATED)
+    def set_thread_number(self, number):
+        """
+        Set the number of threads for the alignment run.
+
+        Parameters
+        ----------
+        number : int, optional
+            The number of threads.
+        """
+        self._n_threads = number
+
+    @requires_state(AppState.CREATED)
+    def use_super5(self):
+        """
+        Use the *Super5* algorithm for the alignment run.
+        """
+        self._mode = "super5"
+
+    def run(self):
+        args = [
+            f"-{self._mode}",
+            self.get_input_file_path(),
+            "-output", self.get_output_file_path(),
+        ]
+        if self.get_seqtype() == "protein":
+            args += ["-amino"]
+        else:
+            args += ["-nt"]
+        if self._n_threads is not None:
+             args += ["-threads", str(self._n_threads)]
+        if self._consiters is not None:
+             args += ["-consiters", str(self._consiters)]
+        if self._refineiters is not None:
+             args += ["-refineiters", str(self._refineiters)]
+        self.set_arguments(args)
+        super().run()
+
+    def clean_up(self):
+        super().clean_up()
+
+    @staticmethod
+    def supports_nucleotide():
+        return True
+
+    @staticmethod
+    def supports_protein():
+        return True
+
+    @staticmethod
+    def supports_custom_nucleotide_matrix():
+        return False
+
+    @staticmethod
+    def supports_custom_protein_matrix():
+        return False
+
+    @classmethod
+    def align(cls, sequences, bin_path=None):
+        """
+        Perform a multiple sequence alignment.
+        
+        This is a convenience function, that wraps the :class:`Muscle5App`
+        execution.
+        
+        Parameters
+        ----------
+        sequences : iterable object of Sequence
+            The sequences to be aligned
+        bin_path : str, optional
+            Path of the MSA software binary. By default, the default path
+            will be used.
+        
+        Returns
+        -------
+        alignment : Alignment
+            The global multiple sequence alignment.
+        """
+        app = cls(sequences, bin_path)
+        app.start()
+        app.join()
+        return app.get_alignment()