snplib 1.0.0__py3-none-any.whl

format/tests/test_plink_ped.py

@@ -0,0 +1,136 @@
+#!/usr/bin/env python
+# coding: utf-8
+__author__ = "Igor Loschinin (igor.loschinin@gmail.com)"
+
+from . import DIR_FILES
+from .. import make_ped
+
+import pytest
+import pandas as pd
+
+
+@pytest.fixture
+def data_ped(request) -> pd.DataFrame | None:
+	return pd.read_pickle(DIR_FILES / f"fplink/ped/{request.param}")
+
+
+class TestPlinkFormatPed(object):
+
+	@pytest.mark.parametrize("data_ped", ["file.pl"], indirect=True)
+	def test_ped_true(self, data_ped: pd.DataFrame) -> None:
+		assert not make_ped(
+			data_ped,
+			"SAMPLE_ID",
+			"SNP",
+			fid_col="SAMPLE_ID"
+		).empty
+
+		assert not make_ped(
+			data_ped,
+			"SAMPLE_ID",
+			"SNP"
+		).empty
+
+	def test_ped_empty(self) -> None:
+		assert make_ped(
+			pd.DataFrame(columns=["SAMPLE_ID", "SNP"]),
+			"SAMPLE_ID",
+			"SNP"
+		).empty
+
+		assert make_ped(
+			pd.DataFrame(columns=["SAMPLE_ID", "SNP"]),
+			"SAMPLE_ID",
+			"SNP",
+			fid_col="SAMPLE_ID"
+		).empty
+
+	@pytest.mark.parametrize("data_ped", ["file.pl"], indirect=True)
+	def test_ped_raise_columns(self, data_ped: pd.DataFrame) -> None:
+		# SID_COL
+		with pytest.raises(
+				KeyError, match="Data has not in name columns!"
+		):
+			make_ped(
+				data=data_ped,
+				sid_col="SAMPLE_ID1",
+				fid_col="SAMPLE_ID",
+				snp_col="SNP"
+			)
+
+		# SNP_COL
+		with pytest.raises(
+				KeyError, match="Data has not in name columns!"
+		):
+			make_ped(
+				data_ped,
+				"SAMPLE_ID",
+				"SNP1",
+				fid_col="SAMPLE_ID"
+			)
+
+		# FID_COL
+		with pytest.raises(
+				KeyError, match="Data has not in name columns SAMPLE_ID1!"
+		):
+			make_ped(
+				data_ped,
+				"SAMPLE_ID",
+				"SNP",
+				fid_col="SAMPLE_ID1"
+			)
+
+	@pytest.mark.parametrize("data_ped", ["file2.pl"], indirect=True)
+	def test_ped_raises_underscope_sid(self, data_ped: pd.DataFrame) -> None:
+
+		# SID_COL
+		with pytest.raises(
+				Exception,
+				match="Replace in 'Sample ID' columns '_' on another a simbols"
+		):
+			res = make_ped(
+				data_ped,
+				"SAMPLE_ID",
+				"SNP"
+			)
+
+	@pytest.mark.parametrize("data_ped", ["file3.pl"], indirect=True)
+	def test_ped_raises_underscope_fid(self, data_ped: pd.DataFrame) -> None:
+
+		# FID_COL
+		with pytest.raises(
+				Exception,
+				match="Replace in 'Family ID' columns '_' on another a simbols"
+		):
+			res = make_ped(
+				data_ped,
+				"SAMPLE_ID",
+				"SNP",
+				fid_col="FAMILY_ID"
+			)
+
+	@pytest.mark.parametrize("data_ped", ["file4.pl"], indirect=True)
+	def test_ped_check_data(self, data_ped: pd.DataFrame) -> None:
+		res = make_ped(
+			data_ped,
+			"SAMPLE_ID",
+			"SNP",
+			fid_col="FAMILY_ID",
+			father_col="father",
+			mother_col="mother",
+			sex_col="sex"
+		)
+
+		res2 = make_ped(
+			data_ped,
+			"SAMPLE_ID",
+			"SNP",
+			fid_col="FAMILY_ID",
+		)
+
+		assert all(res.father.values == list('1234'))
+		assert all(res.mother.values == list('5678'))
+		assert all(res.sex.values == list('1210'))
+		assert all(res2.father.values == list('0000'))
+		assert all(res2.mother.values == list('0000'))
+		assert all(res2.sex.values == list('0000'))

format/tests/test_snp.py

@@ -0,0 +1,128 @@
+#!/usr/bin/env python
+# coding: utf-8
+__author__ = "Igor Loschinin (igor.loschinin@gmail.com)"
+
+from . import DIR_FILES
+from .. import Snp
+
+import pytest
+import pandas as pd
+
+
+@pytest.fixture
+def data_fr(request) -> pd.DataFrame:
+	return pd.read_csv(DIR_FILES / f"fsnp/{request.param}", sep="\t")
+
+
+@pytest.fixture
+def obj_snp(request) -> Snp:
+	return Snp(fmt=request.param)
+
+
+class TestSNP(object):
+
+	@pytest.mark.parametrize(
+		"obj_snp, data_fr", [("uga", 'file1.txt')], indirect=True
+	)
+	def test_snp_process_uga_true(
+			self, obj_snp: Snp, data_fr: pd.DataFrame
+	) -> None:
+
+		obj_snp.process(data_fr)
+		assert obj_snp.data is not None and not obj_snp.data.empty
+		assert obj_snp.data.SNP.isin([
+			'02011015010000500', '01110152120222512'
+		]).all()
+
+	@pytest.mark.parametrize("obj_snp", ["uga"], indirect=True)
+	def test_snp_process_uga_empty(self, obj_snp: Snp) -> None:
+
+		obj_snp.process(pd.DataFrame(columns=[
+			'SNP Name', 'Sample ID', 'Allele1 - AB', 'Allele2 - AB',
+			'GC Score', 'GT Score'
+		]))
+		assert obj_snp.data is not None and obj_snp.data.empty
+
+	@pytest.mark.parametrize(
+		"obj_snp, data_fr",
+		[("uga", 'file1.txt'), (None, 'file1.txt')],
+		indirect=True
+	)
+	def test_snp_process_raises(
+			self, obj_snp: Snp, data_fr: pd.DataFrame
+	) -> None:
+
+		with pytest.raises(KeyError):
+			obj_snp.process(pd.DataFrame(columns=[
+				'SNP Name1', 'Sample ID1', 'Allele1 - AB', 'Allele2 - AB',
+				'GC Score', 'GT Score'
+			]))
+
+		assert obj_snp.data is None
+
+	@pytest.mark.parametrize(
+		"obj_snp, data_fr", [(None, 'file1.txt')], indirect=True
+	)
+	def test_snp_process_df(
+			self, obj_snp: Snp, data_fr: pd.DataFrame
+	) -> None:
+
+		obj_snp.process(data_fr)
+		assert obj_snp.data is not None and not obj_snp.data.empty
+
+	@pytest.mark.parametrize("obj_snp", [None], indirect=True)
+	def test_snp_process_df_empty(self, obj_snp: Snp) -> None:
+
+		obj_snp.process(pd.DataFrame(columns=[
+			'SNP Name', 'Sample ID', 'Allele1 - AB', 'Allele2 - AB',
+			'GC Score', 'GT Score'
+		]))
+		assert obj_snp.data is not None and obj_snp.data.empty
+
+	@pytest.mark.parametrize(
+		"obj_snp, data_fr", [("uga", 'file1.txt')], indirect=True
+	)
+	def test_snp_to_file_uga1(
+			self, obj_snp: Snp, data_fr: pd.DataFrame, tmp_path
+	) -> None:
+		"""
+		The name sample_id is one length
+		"""
+
+		_dir_sub = tmp_path / "sub"
+		_dir_sub.mkdir()
+		_file_save = _dir_sub / "data_snp.csv"
+
+		obj_snp.process(data_fr)
+		assert obj_snp.data is not None and not obj_snp.data.empty
+
+		obj_snp.to_file(_file_save)
+		assert _file_save.is_file() and _file_save.exists()
+		assert (
+			_file_save.read_text() ==
+			"14814 02011015010000500\n14815 01110152120222512\n"
+		)
+
+	@pytest.mark.parametrize(
+		"obj_snp, data_fr", [("uga", 'file2.txt')], indirect=True
+	)
+	def test_snp_to_file_uga2(
+			self, obj_snp: Snp, data_fr: pd.DataFrame, tmp_path
+	) -> None:
+		"""
+		The name sample_id of different length
+		"""
+
+		_dir_sub = tmp_path / "sub"
+		_dir_sub.mkdir()
+		_file_save = _dir_sub / "data_snp.csv"
+
+		obj_snp.process(data_fr)
+		assert obj_snp.data is not None and not obj_snp.data.empty
+
+		obj_snp.to_file(_file_save)
+		assert _file_save.is_file() and _file_save.exists()
+		assert (
+			_file_save.read_text() ==
+			"14814qwert 02011015010000500\n14815      01110152120222512\n"
+		)

parentage/__init__.py

@@ -0,0 +1,15 @@
+#!/usr/bin/env python
+# coding: utf-8
+__author__ = "Igor Loschinin (igor.loschinin@gmail.com)"
+
+from ._discov import Discovery
+from ._verif import Verification
+from ._isagmark import isag_verif, isag_disc
+
+
+__all__ = [
+	"Discovery",
+	"Verification",
+	"isag_disc",
+	"isag_verif"
+]

parentage/_discov.py

@@ -0,0 +1,102 @@
+#!/usr/bin/env python
+# coding: utf-8
+__author__ = "Igor Loschinin (igor.loschinin@gmail.com)"
+
+import pandas as pd
+
+"""
+Search for paternity according to ICAR recommendations
+https://www.icar.org/Documents/GenoEx/ICAR%20Guidelines%20for%20Parentage%20Verification%20and%20Parentage%20Discovery%20based%20on%20SNP.pdf
+"""
+
+
+class Discovery(object):
+    """ Search for paternity according to ICAR recommendations
+
+    :argument isag_markers: Fixed sample of markers to confirm paternity.
+    """
+
+    def __init__(
+            self, isag_markers: pd.Series | list | set | None = None
+    ) -> None:
+        self.__isag_markers = isag_markers
+
+        self.__num_conflicts = None  # Number of conflicts
+        self.__perc_conflicts = None
+
+    @property
+    def status(self) -> None | str:
+        """ The status of each parent discovered. """
+
+        if self.__perc_conflicts is not None:
+            if 0 <= self.__perc_conflicts < 1:
+                return 'Discovered'
+            elif 1 < self.__perc_conflicts < 3:
+                return 'Doubtful'
+            elif self.__perc_conflicts >= 3:
+                return 'Excluded'
+            else:
+                return None
+
+    @property
+    def num_conflicts(self) -> None | int:
+        return self.__num_conflicts
+
+    @property
+    def perc_conflicts(self) -> None | float:
+        return self.__perc_conflicts
+
+    def search_parent(
+            self,
+            data: pd.DataFrame,
+            descendant: str,
+            parents: str,
+            snp_name_col: str
+    ) -> None:
+        """ Search for paternity.
+
+        :param data: SNP data for descendant and parent.
+        :param descendant: Columns name of the descendant in the data.
+        :param parents: Columns name or list name of the parents in the data.
+        :param snp_name_col: SNP columns name is data.
+        """
+
+        if self.__isag_markers is None:
+            raise ValueError("Error. No array of snp names to verify")
+
+        sample_by_markers = data.loc[
+            data[snp_name_col].isin(self.__isag_markers),
+            [snp_name_col, descendant, parents]
+        ]
+
+        # Filtering 5s from a descendent
+        desc_marks = sample_by_markers.loc[
+            sample_by_markers[descendant] != 5, [snp_name_col, descendant]
+        ]
+
+        # According to ICAR, the number of available markers must be
+        # above 450
+        if len(desc_marks) < 450:
+            raise Exception("Calf call rate is low.")
+
+        # Common after filtering markers of potential ancestors
+        sample_parents = sample_by_markers.loc[
+            sample_by_markers[snp_name_col].isin(desc_marks[snp_name_col]),
+            parents
+        ]
+
+        # Number of available markers in potential ancestors
+        prob_parents_same_n_markers = (sample_parents < 5).sum()
+
+        # number of conflicts
+        self.__num_conflicts = (
+            abs(sample_parents.sub(desc_marks[descendant], axis=0)) == 2
+        ).sum()
+
+        # Percentage of conflicts
+        self.__perc_conflicts = (
+            (self.__num_conflicts / prob_parents_same_n_markers) * 100
+        ).round(2)
+
+    def __status_define(self) -> None:
+        ...

parentage/_isagmark.py

@@ -0,0 +1,15 @@
+#!/usr/bin/env python
+# coding: utf-8
+__author__ = "Igor Loschinin (igor.loschinin@gmail.com)"
+
+from pathlib import Path
+
+import pandas as pd
+
+
+def isag_disc() -> pd.DataFrame:
+	return pd.read_pickle(Path(__file__).parent.joinpath("isag_disc.pl"))
+
+
+def isag_verif() -> pd.DataFrame:
+	return pd.read_pickle(Path(__file__).parent.joinpath("isag_verif.pl"))

parentage/_verif.py

@@ -0,0 +1,91 @@
+#!/usr/bin/env python
+# coding: utf-8
+__author__ = "Igor Loschinin (igor.loschinin@gmail.com)"
+
+import numpy as np
+import pandas as pd
+
+
+"""
+https://www.icar.org/Documents/GenoEx/ICAR%20Guidelines%20for%20Parentage%20Verification%20and%20Parentage%20Discovery%20based%20on%20SNP.pdf    
+"""
+
+
+class Verification(object):
+    """
+    Verification of paternity according to ICAR recommendations.
+
+    :argument isag_marks: Fixed sample of markers to confirm paternity.
+    """
+
+    def __init__(
+            self, isag_marks: pd.Series | list | set | None = None
+    ) -> None:
+        self.__isag_marks = isag_marks
+
+        # The minimum number of SNP available in the profile
+        # of each animal and potential parent must be scaled (i.e.: 95%
+        # truncated down)
+        self.__min_num_snp = 0.95
+        self.__num_conflicts = None  # Number of conflicts
+
+    @property
+    def status(self) -> None | str:
+        if self.__num_conflicts is not None:
+            if self.__num_conflicts <= 2:
+                return 'Accept'
+            elif 3 <= self.__num_conflicts <= 5:
+                return 'Doubtful'
+            elif self.__num_conflicts > 5:
+                return 'Excluded'
+            else:
+                return None
+
+    @property
+    def num_conflicts(self) -> None | int:
+        return self.__num_conflicts
+
+    def check_on(
+            self,
+            data: pd.DataFrame,
+            descendant: str,
+            parent: str,
+            snp_name_col: str
+    ) -> None:
+        """ Verification of paternity according to ICAR recommendations.
+
+        :param data: SNP data for descendant and parent.
+        :param descendant: Columns name of the descendant in the data.
+        :param parent: Columns name of the parent in the data.
+        :param snp_name_col: SNP column name in data.
+        """
+
+        if self.__isag_marks is None:
+            raise ValueError('Error. No array of snp names to verify')
+
+        num_isag_mark = len(self.__isag_marks)
+        min_num_comm_snp = int(num_isag_mark - (2 * (num_isag_mark * 0.05)))
+
+        sample_mark = data.loc[
+            data[snp_name_col].isin(self.__isag_marks), [descendant, parent]
+        ]
+
+        # The number of markers is not 5ok
+        desc_n_markers = (sample_mark[descendant] < 5).sum()
+        parent_n_markers = (sample_mark[parent] < 5).sum()
+
+        # According to ICAR, the number of markers not 5ok should be more
+        # than 95%
+        if (desc_n_markers < num_isag_mark * self.__min_num_snp) and \
+                (parent_n_markers < num_isag_mark * self.__min_num_snp):
+            raise Exception('Calf and parent have low call rate')
+
+        comm_snp_no_missing = sample_mark.replace(5, np.nan).dropna()
+        num_comm_markers = len(comm_snp_no_missing)
+
+        if num_comm_markers < min_num_comm_snp:
+            raise Exception('Pair call rate is low')
+
+        self.__num_conflicts = (abs(
+            comm_snp_no_missing[descendant] - comm_snp_no_missing[parent]
+        ) == 2).sum()

parentage/tests/__init__.py

@@ -0,0 +1,7 @@
+#!/usr/bin/env python
+# coding: utf-8
+__author__ = "Igor Loschinin (igor.loschinin@gmail.com)"
+
+from pathlib import Path
+
+DIR_DATA = Path(__file__).parent.joinpath("data")

parentage/tests/test_discov.py

@@ -0,0 +1,164 @@
+#!/usr/bin/env python
+# coding: utf-8
+__author__ = "Igor Loschinin (igor.loschinin@gmail.com)"
+
+from . import DIR_DATA
+from .. import Discovery, isag_disc
+
+import pytest
+import pandas as pd
+
+
+@pytest.fixture
+def data() -> pd.DataFrame:
+	return pd.read_csv(DIR_DATA / "parentage_test_disc.csv", sep=" ")
+
+
+@pytest.fixture
+def obj_discovery() -> Discovery:
+	return Discovery(isag_markers=isag_disc().markers)
+
+
+class TestDiscovery(object):
+
+	def test_search_parent_successfully(
+		self, data: pd.DataFrame, obj_discovery: Discovery
+	) -> None:
+
+		assert obj_discovery.search_parent(
+			data=data,
+			descendant="BY000041988163",
+			parents="EE10512586",
+			snp_name_col="SNP_Name"
+		) is None
+		assert obj_discovery.num_conflicts == 77
+		assert obj_discovery.status == "Excluded"
+		assert obj_discovery.perc_conflicts == 14.86
+
+	def test_search_parent_1(self, data: pd.DataFrame) -> None:
+		"""
+		An exception is thrown for the absence of data with isag markers
+		"""
+		obj_discovery = Discovery()
+
+		with pytest.raises(
+			ValueError, match="Error. No array of snp names to verify"
+		):
+			obj_discovery.search_parent(
+				data=data,
+				descendant="BY000041988163",
+				parents="EE10512586",
+				snp_name_col="SNP_Name"
+			)
+		assert obj_discovery.status is None
+		assert obj_discovery.num_conflicts is None
+		assert obj_discovery.perc_conflicts is None
+
+	def test_search_parent_2(
+		self, data: pd.DataFrame, obj_discovery: Discovery
+	) -> None:
+		"""
+		Exception when the number of markers required to confirm paternity is
+		less than the established value.
+		"""
+
+		with pytest.raises(
+			Exception, match="Calf call rate is low."
+		):
+			obj_discovery.search_parent(
+				data=data[:-100],
+				descendant="BY000041988163",
+				parents="EE10512586",
+				snp_name_col="SNP_Name"
+			)
+		assert obj_discovery.status is None
+		assert obj_discovery.num_conflicts is None
+		assert obj_discovery.perc_conflicts is None
+
+	def test_search_parent_3(
+		self, data: pd.DataFrame, obj_discovery: Discovery
+	) -> None:
+		"""
+		Test if the transmitted animal names are not in the dataframe.
+		"""
+
+		# For descendant
+		with pytest.raises(KeyError):
+			obj_discovery.search_parent(
+				data=data,
+				descendant="BY00004198816",
+				parents="EE10512586",
+				snp_name_col="SNP_Name"
+			)
+		assert obj_discovery.status is None
+		assert obj_discovery.num_conflicts is None
+		assert obj_discovery.perc_conflicts is None
+
+		# For parents
+		with pytest.raises(KeyError):
+			obj_discovery.search_parent(
+				data=data,
+				descendant="BY000041988163",
+				parents="EE105125864",
+				snp_name_col="SNP_Name"
+			)
+		assert obj_discovery.status is None
+		assert obj_discovery.num_conflicts is None
+		assert obj_discovery.perc_conflicts is None
+
+	def test_search_parent_4(
+		self, data: pd.DataFrame, obj_discovery: Discovery
+	) -> None:
+		"""
+		Test when all snp data is not read - equal to 5.
+		"""
+		data[["BY000041988163", "EE10512586"]] = 5
+
+		with pytest.raises(
+			Exception, match="Calf call rate is low."
+		):
+			obj_discovery.search_parent(
+				data=data,
+				descendant="BY000041988163",
+				parents="EE10512586",
+				snp_name_col="SNP_Name"
+			)
+		assert obj_discovery.status is None
+		assert obj_discovery.num_conflicts is None
+		assert obj_discovery.perc_conflicts is None
+
+	def test_search_parent_5(
+			self, data: pd.DataFrame, obj_discovery: Discovery
+	) -> None:
+		"""
+		Test when there is a complete match.
+		"""
+		data[["BY000041988163", "EE10512586"]] = 2
+
+		obj_discovery.search_parent(
+			data=data,
+			descendant="BY000041988163",
+			parents="EE10512586",
+			snp_name_col="SNP_Name"
+		)
+		assert obj_discovery.status == "Discovered"
+		assert obj_discovery.num_conflicts == 0
+		assert obj_discovery.perc_conflicts == 0.0
+
+	def test_search_parent_6(
+			self, data: pd.DataFrame, obj_discovery: Discovery
+	) -> None:
+		"""
+		Partial match test.
+		"""
+		data.loc[202:, "EE10512586"] = 1
+
+		obj_discovery.search_parent(
+			data=data,
+			descendant="BY000041988163",
+			parents="EE10512586",
+			snp_name_col="SNP_Name"
+		)
+		assert obj_discovery.status == "Doubtful"
+		assert obj_discovery.num_conflicts == 14
+		assert obj_discovery.perc_conflicts == 2.70