Datasets:

introvoyz042
/

haddock3

License:

haddock3 / data /tests /test_libclust.py

Migrated from GitHub

5631464 verified about 1 month ago

5.24 kB

"""Test the libclust library."""

import os
import random
import tempfile
from pathlib import Path
from typing import cast

import numpy as np
import pytest

from haddock.libs.libclust import (
    MAX_NB_ENTRY_HTML_MATRIX,
    add_cluster_info,
    plot_cluster_matrix,
    write_structure_list,
)
from haddock.libs.libontology import PDBFile

from . import golden_data


@pytest.fixture
def protprot_input_models():
    """Prot-prot input."""
    return [
        PDBFile(
            Path(golden_data, "protprot_complex_1.pdb"), path=golden_data, score=-42
        ),
        PDBFile(
            Path(golden_data, "protprot_complex_2.pdb"), path=golden_data, score=-17
        ),
    ]


@pytest.fixture
def big_distance_matrix_data():
    """Big distance matrix data."""
    bigmatrix = []
    for i in range(1, MAX_NB_ENTRY_HTML_MATRIX + 1):
        for j in range(i + 1, MAX_NB_ENTRY_HTML_MATRIX + 2):
            bigmatrix.append(f"{i}\t{j}\t{random.random():.2f}")
    return bigmatrix


@pytest.fixture
def small_distance_matrix_data():
    """Distance matrix data."""
    small = 10
    matrix = []
    for i in range(1, small):
        for j in range(i + 1, small + 1):
            matrix.append(f"{i}\t{j}\t{random.random():.2f}")
    return matrix


def test_write_structure_list(protprot_input_models):
    """Test write_structure_list function."""
    # fake clustering
    clustered_models = [protprot_input_models[0]]
    clustered_models[0].clt_id = 1
    cl_fname = "clustfcc.tsv"
    write_structure_list(protprot_input_models, clustered_models, cl_fname)
    observed_file_content = open(cl_fname, "r").read()
    expected_file_content = (
        f"rank\tmodel_name\tscore\tcluster_id{os.linesep}"
        f"1\tprotprot_complex_1.pdb\t-42.00\t1{os.linesep}"
        f"2\tprotprot_complex_2.pdb\t-17.00\t-{os.linesep}"
        f"{os.linesep}"
    )
    assert observed_file_content == expected_file_content
    os.unlink(cl_fname)


def test_plot_cluster_matrix_big(big_distance_matrix_data):
    """Test test_plot_cluster_matrix_big function with big matrix."""
    with tempfile.TemporaryDirectory(dir=".") as tmpdir:
        # Write matrix
        matrix_path = f"{tmpdir}bigmatrix.mat"
        with open(matrix_path, "w") as f:
            f.write("\n".join(big_distance_matrix_data))
        # Run function
        figure_path = plot_cluster_matrix(
            matrix_path,
            list(range(MAX_NB_ENTRY_HTML_MATRIX + 1)),
            [str(i + 1) for i in range(MAX_NB_ENTRY_HTML_MATRIX + 1)],
            dttype="random",
            diag_fill=0.5,
            color_scale="Blues",
            reverse=False,
            output_fname="clust_matrix_test_big",
        )
        # Check output
        assert figure_path is None
        assert not Path("clust_matrix_test_big.html").exists()
        Path(matrix_path).unlink(missing_ok=False)


def test_plot_cluster_matrix(small_distance_matrix_data):
    """Test test_plot_cluster_matrix_big function with small matrix."""
    with tempfile.TemporaryDirectory(dir=".") as tmpdir:
        # Write matrix
        matrix_path = f"{tmpdir}smallmatrix.mat"
        with open(matrix_path, "w") as f:
            f.write("\n".join(small_distance_matrix_data))
        # Run function
        figure_path = plot_cluster_matrix(
            matrix_path,
            list(range(10)),
            [str(i + 1) for i in range(10)],
            dttype="random",
            diag_fill=0.5,
            color_scale="Blues",
            reverse=True,
            output_fname="clust_matrix_test_small",
        )
        # Check output
        assert os.path.exists(figure_path)
        assert Path(figure_path).stat().st_size != 0
        assert Path(figure_path).suffix == ".html"
        Path(figure_path).unlink(missing_ok=False)
        Path(matrix_path).unlink(missing_ok=False)


def test_add_cluster_info(protprot_input_models):
    # first 5 -> pdbs_cluster_2 (worse cluster), last 5 -> pdbs_cluster_1 (better cluster)
    scores = [+10.0, +20.0, +30.0, +40.0, +50.0, -410.0, -420.0, -430.0, -440.0, -450.0]
    input_pdbs = [
        PDBFile(file_name=f"{i}.pdb", score=score) for i, score in enumerate(scores)
    ]

    # split so we get two sets of models
    mid = len(input_pdbs) // 2
    pdbs_cluster_1 = input_pdbs[mid:]
    avg_1 = float(np.average(scores[mid:]))
    pdbs_cluster_2 = input_pdbs[:mid]
    avg_2 = float(np.average(scores[:mid]))

    # create cluster information
    # [(cluster_id, average-score)]
    input_l = [(1, avg_1), (2, avg_2)]

    # Make sure the PDBFiles are not created with any cluster information
    for p in input_pdbs:
        assert p.clt_id is None
        assert p.clt_rank is None
        assert p.clt_model_rank is None

    observed = add_cluster_info(input_l, clt_dic={1: pdbs_cluster_1, 2: pdbs_cluster_2})

    # assert that the cluster information was added to the models
    for p in observed:
        assert p.clt_id is not None
        assert p.clt_rank is not None
        assert p.clt_model_rank is not None

    # assert the order has changed
    assert observed[0] is not input_pdbs[0]

    # check they were properly sorted
    assert cast(float, observed[0].score) < cast(float, observed[-1].score)

	"""Test the libclust library."""

	import os
	import random
	import tempfile
	from pathlib import Path
	from typing import cast

	import numpy as np
	import pytest

	from haddock.libs.libclust import (
	MAX_NB_ENTRY_HTML_MATRIX,
	add_cluster_info,
	plot_cluster_matrix,
	write_structure_list,
	)
	from haddock.libs.libontology import PDBFile

	from . import golden_data


	@pytest.fixture
	def protprot_input_models():
	"""Prot-prot input."""
	return [
	PDBFile(
	Path(golden_data, "protprot_complex_1.pdb"), path=golden_data, score=-42
	),
	PDBFile(
	Path(golden_data, "protprot_complex_2.pdb"), path=golden_data, score=-17
	),
	]


	@pytest.fixture
	def big_distance_matrix_data():
	"""Big distance matrix data."""
	bigmatrix = []
	for i in range(1, MAX_NB_ENTRY_HTML_MATRIX + 1):
	for j in range(i + 1, MAX_NB_ENTRY_HTML_MATRIX + 2):
	bigmatrix.append(f"{i}\t{j}\t{random.random():.2f}")
	return bigmatrix


	@pytest.fixture
	def small_distance_matrix_data():
	"""Distance matrix data."""
	small = 10
	matrix = []
	for i in range(1, small):
	for j in range(i + 1, small + 1):
	matrix.append(f"{i}\t{j}\t{random.random():.2f}")
	return matrix


	def test_write_structure_list(protprot_input_models):
	"""Test write_structure_list function."""
	# fake clustering
	clustered_models = [protprot_input_models[0]]
	clustered_models[0].clt_id = 1
	cl_fname = "clustfcc.tsv"
	write_structure_list(protprot_input_models, clustered_models, cl_fname)
	observed_file_content = open(cl_fname, "r").read()
	expected_file_content = (
	f"rank\tmodel_name\tscore\tcluster_id{os.linesep}"
	f"1\tprotprot_complex_1.pdb\t-42.00\t1{os.linesep}"
	f"2\tprotprot_complex_2.pdb\t-17.00\t-{os.linesep}"
	f"{os.linesep}"
	)
	assert observed_file_content == expected_file_content
	os.unlink(cl_fname)


	def test_plot_cluster_matrix_big(big_distance_matrix_data):
	"""Test test_plot_cluster_matrix_big function with big matrix."""
	with tempfile.TemporaryDirectory(dir=".") as tmpdir:
	# Write matrix
	matrix_path = f"{tmpdir}bigmatrix.mat"
	with open(matrix_path, "w") as f:
	f.write("\n".join(big_distance_matrix_data))
	# Run function
	figure_path = plot_cluster_matrix(
	matrix_path,
	list(range(MAX_NB_ENTRY_HTML_MATRIX + 1)),
	[str(i + 1) for i in range(MAX_NB_ENTRY_HTML_MATRIX + 1)],
	dttype="random",
	diag_fill=0.5,
	color_scale="Blues",
	reverse=False,
	output_fname="clust_matrix_test_big",
	)
	# Check output
	assert figure_path is None
	assert not Path("clust_matrix_test_big.html").exists()
	Path(matrix_path).unlink(missing_ok=False)


	def test_plot_cluster_matrix(small_distance_matrix_data):
	"""Test test_plot_cluster_matrix_big function with small matrix."""
	with tempfile.TemporaryDirectory(dir=".") as tmpdir:
	# Write matrix
	matrix_path = f"{tmpdir}smallmatrix.mat"
	with open(matrix_path, "w") as f:
	f.write("\n".join(small_distance_matrix_data))
	# Run function
	figure_path = plot_cluster_matrix(
	matrix_path,
	list(range(10)),
	[str(i + 1) for i in range(10)],
	dttype="random",
	diag_fill=0.5,
	color_scale="Blues",
	reverse=True,
	output_fname="clust_matrix_test_small",
	)
	# Check output
	assert os.path.exists(figure_path)
	assert Path(figure_path).stat().st_size != 0
	assert Path(figure_path).suffix == ".html"
	Path(figure_path).unlink(missing_ok=False)
	Path(matrix_path).unlink(missing_ok=False)


	def test_add_cluster_info(protprot_input_models):
	# first 5 -> pdbs_cluster_2 (worse cluster), last 5 -> pdbs_cluster_1 (better cluster)
	scores = [+10.0, +20.0, +30.0, +40.0, +50.0, -410.0, -420.0, -430.0, -440.0, -450.0]
	input_pdbs = [
	PDBFile(file_name=f"{i}.pdb", score=score) for i, score in enumerate(scores)
	]

	# split so we get two sets of models
	mid = len(input_pdbs) // 2
	pdbs_cluster_1 = input_pdbs[mid:]
	avg_1 = float(np.average(scores[mid:]))
	pdbs_cluster_2 = input_pdbs[:mid]
	avg_2 = float(np.average(scores[:mid]))

	# create cluster information
	# [(cluster_id, average-score)]
	input_l = [(1, avg_1), (2, avg_2)]

	# Make sure the PDBFiles are not created with any cluster information
	for p in input_pdbs:
	assert p.clt_id is None
	assert p.clt_rank is None
	assert p.clt_model_rank is None

	observed = add_cluster_info(input_l, clt_dic={1: pdbs_cluster_1, 2: pdbs_cluster_2})

	# assert that the cluster information was added to the models
	for p in observed:
	assert p.clt_id is not None
	assert p.clt_rank is not None
	assert p.clt_model_rank is not None

	# assert the order has changed
	assert observed[0] is not input_pdbs[0]

	# check they were properly sorted
	assert cast(float, observed[0].score) < cast(float, observed[-1].score)