ibdpainting 0.1__tar.gz

ibdpainting-0.1/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Tom Ellis
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

ibdpainting-0.1/PKG-INFO

@@ -0,0 +1,18 @@
+Metadata-Version: 2.1
+Name: ibdpainting
+Version: 0.1
+Summary: Identify parents of a crossed individual by comparing identity in windows across their genomes
+Home-page: https://github.com/ellisztamas/ibdpainting
+Author: Tom Ellis
+Author-email: thomas.ellis@gmi.oeaw.ac.at
+License: MIT
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy
+Requires-Dist: pandas
+Requires-Dist: plotly
+Requires-Dist: h5py
+Requires-Dist: scikit-allel
+
+# ibdpainting
+Identify parents of a crossed individual by comparing identity in windows across their genomes

ibdpainting-0.1/README.md

@@ -0,0 +1,2 @@
+# ibdpainting
+Identify parents of a crossed individual by comparing identity in windows across their genomes

ibdpainting-0.1/ibdpainting/__init__.py

@@ -0,0 +1,8 @@
+"""Top-level package for methlab."""
+
+__author__ = """Tom Ellis"""
+__email__ = 'thomas.ellis@gmi.oeaw.ac.at'
+__version__ = '0.1'
+
+
+from ibdpainting import *

ibdpainting-0.1/ibdpainting/genetic_distance.py

@@ -0,0 +1,127 @@
+import numpy as np
+from warnings import warn
+import numpy.ma as ma
+
+class geneticDistance(object):
+    """
+    A simple class to compare genotype data genetic distances between individuals.  
+
+    Parameters
+    ==========
+    samples: array
+        Vector of length m giving names for each sample.
+    chr: array
+        Vector of length n giving chromosome labels for each SNP.
+    pos: array
+        Vector of length n giving SNP positions. Note that SNP positions are inherited from 
+        skikit allel and give row numbers from the input VCF file rather than
+        base-pair positions on each chromosome.
+    geno: array
+        m x n x 2 array of genotype data where axes index SNPs, individuals, and 
+        homologous chromosomes.
+
+    Attributes
+    ==========
+    samples: array
+        Vector of M sample names. The first sample is the input individual to be
+        compared to the remaining reference individuals.
+    chr: array
+        Vector of chromosome labels. These are imported from the reference panel.
+    pos: array
+        Vector of N SNP positions. These are imported from the reference panel.
+    geno: array
+        NxMx2 array of genotype data, where N is the number of SNPs and M is the
+        number of samples.
+
+    Methods
+    =======
+    split_into_windows
+        Split a geneticDistance object into windows.
+    pairwise_distance
+        Calculate pairwise genetic distance between an input individual and all 
+        reference individuals.
+    
+    """
+    def __init__(self, samples, chr, pos, geno):
+        self.samples = samples
+        self.chr = chr
+        self.pos = pos
+        self.geno = geno
+
+    def split_into_windows(self, window_size: int):
+        """
+        Split a geneticDistance object into windows.
+
+        Splits the geneticDistance object into chromosomes, then into windows on each
+        chromosome. It returns a dictionary of geneticDistance objects for each window.
+
+        Parameters
+        ==========
+        window_size: int
+            Window size in base pairs.
+
+        Returns
+        =======
+        A dictionary of geneticDistance objects with an element for each window.
+        Indexes are in the form "Chr:start-stop".
+        """
+        # Empty dict to store the output
+        list_of_distance_objects = {}
+
+        for chr in np.unique(self.chr):
+            # Boolean array indexing items in this chromosome
+            chr_ix = self.chr == chr
+            
+            # Array of starting positions for each window. 
+            start_positions = np.arange(0, self.pos[chr_ix].max(), window_size)
+            for start in start_positions:
+                stop  = start + window_size
+                # Index positions of SNPs within the current window
+                window_ix = (self.pos[chr_ix] >= start) & (self.pos[chr_ix] < stop)
+                # Create an object for each window.
+                window_name = str(chr) + ":" + str(start) + "-" + str(stop)
+                list_of_distance_objects[window_name] = geneticDistance(
+                        samples = self.samples,
+                        chr  = self.chr[chr_ix][window_ix],
+                        pos  = self.pos[chr_ix][window_ix],
+                        geno = self.geno[chr_ix][window_ix]
+                    )
+        
+        return list_of_distance_objects
+    
+    def pairwise_distance(self, warn_about_missing_data=False):
+        """
+        Calculate pairwise genetic distance between an input individual and all 
+        reference individuals.
+
+        The input individual is always the first in the list of samples. Genetic
+        distance is the number of allelic differences at each locus between each
+        pair, summed over all loci. The calculation is done using masked arrays to
+        account for missing data.
+
+        Returns
+        =======
+        Vector of distances
+
+        """
+        masked_geno = ma.masked_array(self.geno, self.geno < 0)
+
+        # Calculate differences at each locus
+        per_locus_difference = abs(masked_geno.sum(2)[:,[0]] - masked_geno.sum(2)[:,1:]) / 2
+        # Average over loci
+        dxy = per_locus_difference.mean(0)
+
+        if warn_about_missing_data and any(dxy.mask):
+            warn("""
+
+    Pairwise distance could not be calculated for one or more comparisons,
+    probably because there is missing data at all SNPs.
+    The following samples in the reference panel are affected:
+    {}
+
+    """.format(self.samples[1:][dxy.mask])
+            )
+            # Return a vector of -9 t indicate missing data
+            return np.zeros(len(self.samples)-1) -9
+        
+        return dxy.data

ibdpainting-0.1/ibdpainting/ibd_table.py

@@ -0,0 +1,47 @@
+import pandas as pd
+
+from ibdpainting import load_genomes
+
+def ibd_table(input:str, reference:str, sample_name:str, window_size:int):
+    """
+    Compare allele sharing across the genome.
+
+    Calculate genetic distance between a test individual and a panel of
+    reference genomes.
+
+    Parameters
+    ==========
+    input: str
+        Path to a VCF file containing genotype data for one or more samples to
+        test
+    reference: str
+        Path to an HDF5 file containing genotype data for a panel of reference
+        individuals to compare the test individual against.
+    sample_name: str
+        Sample name for the individual to check.
+        This must be present in the samples in the input VCF.
+    window_size: int
+        Window size in base pairs.
+
+    Returns
+    =======
+    DataFrame with a row for each window in the genome and a column for each 
+    sample in the reference panel. Elements show genetic distance between the 
+    test individual and each reference individual in a single window.
+    """
+    genetic_distance = load_genotypes(
+            input = input,
+            reference = reference,
+            sample_name = sample_name
+        )
+    # Divide the genome into windows
+    distances_in_windows = genetic_distance.split_into_windows(window_size)
+
+    # Dataframe with a row for each window across the genome and a column for each sample in the reference panel.
+    distance_array = pd.DataFrame(
+        [ v.pairwise_distance() for v in distances_in_windows.values() ],
+        columns = genetic_distance.samples[1:]
+    )
+    distance_array.insert(0, 'window', distances_in_windows.keys())
+
+    return distance_array

ibdpainting-0.1/ibdpainting/ibs_scores.py

@@ -0,0 +1,36 @@
+import numpy as np
+import pandas as pd
+
+def ibd_scores(ibd_table):
+    """
+    Mean IBD across the genome.
+
+    Calculate mean genetic distance from a test individual to each of a panel of
+    reference samples, ignoring windows where there was only missing data.
+
+    Parameters
+    ==========
+    ibd_table: pd.DataFrame
+        DataFrame with a row for each window in the genome and a column for each 
+        sample in the reference panel. Elements show genetic distance between the 
+        test individual and each reference individual in a single window.
+        This is generated by ibd_table().
+
+    Returns
+    =======
+    A DataFrame with a row for each candidate in the reference panel, and a 
+    column indicating mean genetic distance over windows across the genome.
+    Values closer to zero indicate that the sample is more likely to be a match.
+    """
+    # Coerce missing data to NaN for correct column means.
+    ibd_table = ibd_table.replace(-9,np.NaN)
+
+    # Get column-mean IBD for each candidate, allowing for missing data
+    ibd_scores_for_each_candidate = np.array(
+        [ np.nanmean(ibd_table[col]) for col in ibd_table.keys()[1:] ]
+    )
+    scores = pd.DataFrame({
+        'candidate': ibd_table.keys()[1:],
+        'score' : ibd_scores_for_each_candidate
+    })
+    return scores

ibdpainting-0.1/ibdpainting/load_genomes.py

@@ -0,0 +1,99 @@
+import allel
+import h5py
+import numpy as np
+from ibdpainting import geneticDistance
+
+def load_genotypes(input, reference, sample_name):
+    """
+    Import and merge test and reference data files.
+
+    Import genotype data for one or more input samples and a panel of reference samples
+    to compare to. Subset each so that the markers are really identical. Merge the
+    arrays of genotype calls so that the data for the input appear first on the
+    first axis of the genotype call arrays.
+
+    Parameters
+    ==========
+    input: str
+        Path to a VCF file containing genotype data for one or more samples to check
+    reference: str
+        Path to a HDF5 file containing genotype data for a panel of reference individuals
+        to compare the input indivual against.
+    sample_name: str
+        Sample name for the individual to check. This must be present in the samples
+        in the input VCF.
+
+    Return
+    ======
+    An object of class geneticDistance.
+    """
+    # Read in the data files
+    input_vcf = allel.read_vcf(input, samples=[sample_name])
+    ref_hdf5  = h5py.File(reference, mode="r")
+
+    ref_str_data = {
+        'samples' : [ x.decode('utf-8') for x in ref_hdf5['samples'][:] ],
+        'chr'     : [ x.decode('utf-8') for x in ref_hdf5['variants/CHROM'][:] ]
+    }
+
+    if sample_name not in input_vcf['samples']:
+        raise ValueError("The sample name is not in the list of samples in the input VCF file.")
+    else: 
+        # Find the position of the individual to test
+        sample_ix = np.where(input_vcf['samples'] == sample_name)[0][0]
+        # Join vectors of sample names, with the test individual first
+        new_samples = np.append(
+            input_vcf['samples'][None,sample_ix],
+            ref_str_data['samples']
+            )        
+
+    # Check that contig labels match
+    chr_labels = {
+        'input' : np.unique(input_vcf['variants/CHROM']),
+        'ref'   : np.unique(ref_str_data['chr'])
+    }
+    if len(chr_labels['input']) != len(chr_labels['ref']):
+        raise ValueError(
+            "The number of unique contig labels do not match: the input VCF has {}, but the reference panel has {}.".
+            format( chr_labels['input'], chr_labels['ref'] )
+        )
+    elif any( chr_labels['input'] != chr_labels['ref'] ):
+        raise ValueError(
+            "Contig labels do not match between the input and reference files."
+        )
+    
+    # Make sure we only compare SNPs that are found in both datasets.
+    # Concatenate chromosome labels and SNP positions
+    snp_names = {
+        'input' : [ str(chr) + ":" + str(pos) for chr,pos in zip(input_vcf['variants/CHROM'], input_vcf['variants/POS']) ],
+        'ref'   : [ str(chr) + ":" + str(pos) for chr,pos in zip(ref_str_data['chr'],   ref_hdf5['variants/POS'][:]) ]
+    }
+    # Find the SNP position names that are common to both datasets
+    matching_SNPs_in_both_files = np.intersect1d(
+        snp_names['input'],
+        snp_names['ref']
+        )
+    which_SNPs_to_keep = {
+        "input" : [ x in matching_SNPs_in_both_files for x in snp_names['input'] ],
+        "ref"   : [ x in matching_SNPs_in_both_files for x in snp_names['ref'] ]
+    }
+
+
+    # Append the genotype data for the test individual to the array of the reference panel
+    new_geno = np.concatenate(
+        (input_vcf['calldata/GT'][which_SNPs_to_keep['input'], sample_ix][:, np.newaxis],
+        ref_hdf5['calldata/GT'][which_SNPs_to_keep['ref']]),
+        axis=1
+        )
+    
+    # Define an output before closing the Hdf5 file
+    output = geneticDistance(
+        samples = new_samples,
+        chr = np.array(ref_str_data['chr'])[np.where(which_SNPs_to_keep['ref'])[0]],
+        pos = ref_hdf5['variants/POS'][:][which_SNPs_to_keep['ref']],
+        geno = new_geno
+    )
+
+    ref_hdf5.close()
+
+    return output

ibdpainting-0.1/ibdpainting/plot_ibd_table.py

@@ -0,0 +1,80 @@
+import pandas as pd
+import numpy as np
+
+import plotly.express as px
+
+
+def plot_ibd_table(ibd_table:pd.DataFrame, sample_name:str, expected_match:list=[], max_to_plot=20):
+    """
+    Plot allele sharing across the genome.
+
+
+    Create a interactive line graph showing genetic distance from a test
+    individual to each sample in a panel of reference individuals.
+
+    Parameters
+    ==========
+    ibd_table: pd.DataFrame
+        DataFrame with a row for each window in the genome and a column for each 
+        sample in the reference panel. Elements show genetic distance between the 
+        test individual and each reference individual in a single window.
+        This is generated by ibd_table().
+    sample_name: str
+        Sample name for the individual to check.
+        This must be present in the samples in the input VCF.
+    expected_match: list
+        List of sample names in the reference panel that are expected to be
+        ancestors of the test individual.
+
+    Returns
+    =======
+    Plotly figure object with subplots for each chromosome, showing window
+    position along the x-axis and genetic distance from the test individual to
+    each reference sample on the y-axis. Line colour indicates whether a sample
+    is an expected parent or not. Rolling over the lines shows which sample is
+    which.
+    """
+
+    
+    # Coerce missing data to NaN for correct column means.
+    ibd_table = ibd_table.replace(-9,np.NaN)
+
+    # Identify the candidate names *not* among the top `max_to_plot` columns and remove
+    # If `max_to_plot` is less than the number of candidates.
+    if max_to_plot < ibd_table.shape[1]-1:
+        # Get column-mean IBD for each candidate, allowing for missing data
+        ibd_scores_for_each_candidate = np.array(
+            [ np.nanmean(ibd_table[col]) for col in ibd_table.keys()[1:] ]
+        )
+        # Identify the candidate names *not* among the top `max_to_plot` columns
+        ix = np.argpartition(ibd_scores_for_each_candidate, max_to_plot)[max_to_plot:] # index positions
+        columns_to_drop = ibd_table.keys()[ix+1].to_list() # candidate names
+        ibd_table = ibd_table.drop(columns=columns_to_drop) # drop the candidates
+
+    # Make the table long
+    ibd_table = ibd_table.melt(id_vars=['window'], var_name='candidate', value_name='distance')
+    # Column indicating which candidates should be plotted a different colour.
+    ibd_table['expected'] = ibd_table['candidate'].isin(expected_match)
+    # Split the 'window' column up into separate columns for chromosome, start and stop positions
+    ibd_table[['chr', 'window']] = ibd_table['window'].str.split(":", expand=True)
+    ibd_table[['start', 'stop']] = ibd_table['window'].str.split("-", expand=True)
+    # start and stop positions should be integers for sensible plotting.
+    ibd_table['start'] = ibd_table['start'].astype(int)
+    ibd_table['stop'] = ibd_table['stop'].astype(int)
+    ibd_table['midpoint'] = (ibd_table['start'] + ibd_table['stop']) / 2
+
+    fig = px.line(
+        ibd_table,
+        x="midpoint", y="distance", color="expected",
+        title=sample_name,
+        labels={
+            'midpoint' : 'Position (bp)',
+            'distance' : 'Genetic distance'        
+        },
+        hover_data=['candidate'],
+        category_orders={'expected': [False, True]},
+        facet_row = "chr"
+        )
+    fig.update_traces(mode="markers+lines")
+
+    return fig

ibdpainting-0.1/ibdpainting.egg-info/PKG-INFO

@@ -0,0 +1,18 @@
+Metadata-Version: 2.1
+Name: ibdpainting
+Version: 0.1
+Summary: Identify parents of a crossed individual by comparing identity in windows across their genomes
+Home-page: https://github.com/ellisztamas/ibdpainting
+Author: Tom Ellis
+Author-email: thomas.ellis@gmi.oeaw.ac.at
+License: MIT
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy
+Requires-Dist: pandas
+Requires-Dist: plotly
+Requires-Dist: h5py
+Requires-Dist: scikit-allel
+
+# ibdpainting
+Identify parents of a crossed individual by comparing identity in windows across their genomes

ibdpainting-0.1/ibdpainting.egg-info/SOURCES.txt

@@ -0,0 +1,15 @@
+LICENSE
+README.md
+setup.py
+ibdpainting/__init__.py
+ibdpainting/genetic_distance.py
+ibdpainting/ibd_table.py
+ibdpainting/ibs_scores.py
+ibdpainting/load_genomes.py
+ibdpainting/plot_ibd_table.py
+ibdpainting.egg-info/PKG-INFO
+ibdpainting.egg-info/SOURCES.txt
+ibdpainting.egg-info/dependency_links.txt
+ibdpainting.egg-info/not-zip-safe
+ibdpainting.egg-info/requires.txt
+ibdpainting.egg-info/top_level.txt

ibdpainting-0.1/ibdpainting.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

ibdpainting-0.1/ibdpainting.egg-info/not-zip-safe

@@ -0,0 +1 @@
+

ibdpainting-0.1/ibdpainting.egg-info/requires.txt

@@ -0,0 +1,5 @@
+numpy
+pandas
+plotly
+h5py
+scikit-allel

ibdpainting-0.1/ibdpainting.egg-info/top_level.txt

@@ -0,0 +1 @@
+ibdpainting

ibdpainting-0.1/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

ibdpainting-0.1/setup.py

@@ -0,0 +1,41 @@
+from setuptools import setup
+import codecs
+import os.path
+from pathlib import Path
+
+# Functions to pull the package version from init.py
+def read(rel_path):
+    here = os.path.abspath(os.path.dirname(__file__))
+    with codecs.open(os.path.join(here, rel_path), 'r') as fp:
+        return fp.read()
+
+def get_version(rel_path):
+    for line in read(rel_path).splitlines():
+        if line.startswith('__version__'):
+            delim = '"' if '"' in line else "'"
+            return line.split(delim)[1]
+    else:
+        raise RuntimeError("Unable to find version string.")
+
+# read the contents of your README file
+this_directory = Path(__file__).parent
+long_description = (this_directory / "README.md").read_text()
+
+
+setup(
+    name='ibdpainting',
+    version=get_version("ibdpainting/__init__.py"),
+    description='Identify parents of a crossed individual by comparing identity in windows across their genomes',
+    url='https://github.com/ellisztamas/ibdpainting',
+    long_description=long_description,
+    long_description_content_type='text/markdown',
+    author='Tom Ellis',
+    author_email='thomas.ellis@gmi.oeaw.ac.at',
+    license='MIT',
+    packages=['ibdpainting'],
+    install_requires=[
+        'numpy', 'pandas', 'plotly', 'h5py', 'scikit-allel'
+      ],
+    zip_safe=False
+    )
+