Related Samples Module¶

The usage of the standalone module is shown below:

$ pyGenClean_find_related_samples --help
usage: pyGenClean_find_related_samples [-h] [-v] --bfile FILE [--genome-only]
                                       [--min-nb-snp INT]
                                       [--indep-pairwise STR STR STR]
                                       [--maf FLOAT] [--ibs2-ratio FLOAT]
                                       [--sge] [--sge-walltime TIME]
                                       [--sge-nodes INT INT]
                                       [--line-per-file-for-sge INT]
                                       [--out FILE]

Finds related samples according to IBS values.

optional arguments:
  -h, --help            show this help message and exit
  -v, --version         show program's version number and exit

Input File:
  --bfile FILE          The input file prefix (will find the plink binary
                        files by appending the prefix to the .bim, .bed and
                        .fam files, respectively.)

Options:
  --genome-only         Only create the genome file
  --min-nb-snp INT      The minimum number of markers needed to compute IBS
                        values. [Default: 10000]
  --indep-pairwise STR STR STR
                        Three numbers: window size, window shift and the r2
                        threshold. [default: ['50', '5', '0.1']]
  --maf FLOAT           Restrict to SNPs with MAF >= threshold. [default:
                        0.05]
  --ibs2-ratio FLOAT    The initial IBS2* ratio (the minimum value to show in
                        the plot. [default: 0.8]
  --sge                 Use SGE for parallelization.
  --sge-walltime TIME   The walltime for the job to run on the cluster. Do not
                        use if you are not required to specify a walltime for
                        your jobs on your cluster (e.g. 'qsub
                        -lwalltime=1:0:0' on the cluster).
  --sge-nodes INT INT   The number of nodes and the number of processor per
                        nodes to use (e.g. 'qsub -lnodes=X:ppn=Y' on the
                        cluster, where X is the number of nodes and Y is the
                        number of processor to use. Do not use if you are not
                        required to specify the number of nodes for your jobs
                        on the cluster.
  --line-per-file-for-sge INT
                        The number of line per file for SGE task array.
                        [default: 100]

Output File:
  --out FILE            The prefix of the output files. [default: ibs]

Input Files¶

This module uses PLINK’s binary file format (bed, bim and fam files) for the source data set (the data of interest).

Procedure¶

Here are the steps performed by the module:

Uses Plink to extract markers according to LD.
Checks if there is enough markers after pruning.
Extract markers according to LD.
Runs Plink with the genome option to compute the IBS values.
Finds related individuals and gets values for plotting.
Plots Z1 in function of IBS2 ratio for related individuals.
Plots Z2 in function of IBS2 ratio for related individuals.

Output Files¶

The output files of each of the steps described above are as follow (note that the output prefix shown is the one by default [i.e. ibs]):

One set of PLINK’s result files:
- ibs.pruning_0.1: the results of the pruning process of Plink. The value depends on the option of --indep-pairwise. The markers that are kept are in the file ibs.pruning_0.1.prune.in.
No file created.
One set of PLINK’s binary files:
- ibs.pruned_data: the data sets containing only the marker from the first step (the list is in ibs.pruning_0.1.prune.in).
One set of PLINK’s result files (two if --sge is used):
- ibs.frequency: PLINK’s result files when computing the frequency of each of the pruned markers. This data set will exist only if the option --sge is used.
- ibs.genome: PLINK’s results including IBS values.
One file provided by the pyGenClean.RelatedSamples.find_related_samples and three files provided by pyGenClean.RelatedSamples.merge_related_samples:
- ibs.related_individuals: a subset of the ibs.genome.genome file
  
  containing only samples that are considered to be related. Three columns are appended to the original ibs.genome.genome file: IBS2_ratio (the value that is considered to find related individuals), status (the type of relatedness [e.g. twins]) and code (a numerical code that represent the status). This file is provided by the pyGenClean.RelatedSamples.find_related_samples module.
- ibs.merged_related_individuals: a file aggregating related samples in groups, containing the following columns: index (the group number), FID1 (the family ID of the first sample), IID1 (the individual ID of the first sample), FID2 (the family ID of the second sample), IID2 (the individual ID of the second sample) and status (the type of relatedness between the two samples). This file is provided by the merge_related_samples.
- ibs.chosen_related_individuals: the related individuals that were randomly chosen from each group to be kept in the final data set. This file is provided by the merge_related_samples.
- ibs.discarded_related_individual: the related individuals that needs to be discarded, so that the final data set include only unrelated individuals. This file is provided by the merge_related_samples.
One image file:
- ibs.related_individuals_z1.png: a plot showing the \(Z_1\) value in function of the \(IBS2^*_{ratio}\) for all samples above a certain \(IBS2^*_{ratio}\) (the default threshold is 0.8). See Figure Z1 in function of IBS2 ratio.
One image file:
- ibs.related_individuals_z2.png: a plot showing the \(Z_2\) value in function of the \(IBS2^*_{ratio}\) for all samples above a certain \(IBS2^*_{ratio}\) (the default threshold is 0.8). See Figure Z2 in function of IBS2 ratio.

The Plots¶

The first plot (Z1 in function of IBS2 ratio figure) that is created is \(Z_1\) in function of \(IBS2^*_{ratio}\) (where each point represents a pair of related individuals. The color code comes from the different value of \(Z_0\), \(Z_1\) and \(Z_2\), as described in the pyGenClean.RelatedSamples.find_related_samples.extractRelatedIndividuals() function. In this plot, there are four locations where related samples tend to accumulate (first degree relatives (full sibs), second degree relatives (half-sibs, grand-parent-child or uncle-nephew), parent-child and twins (or duplicated samples). The unknown sample pairs represent possible undetected related individuals.

Z1 in function of IBS2 ratio

The second plot (Z2 in function of IBS2 ratio figure) that is created is \(Z_2\) in function of \(IBS2^*_{ratio}\) (where each point represents a pair of related individuals. It’s just another representation of relatedness of sample pairs, where the location of the “clusters” is different.

Z2 in function of IBS2 ratio

Finding Outliers¶

A standalone script was created in order to regroup related samples in different subset. The usage is as follow:

$ pyGenClean_merge_related_samples --help
usage: pyGenClean_merge_related_samples [-h] [-v] --ibs-related FILE
                                        [--no-status] [--out FILE]

Merges related samples according to IBS.

optional arguments:
  -h, --help          show this help message and exit
  -v, --version       show program's version number and exit

Input File:
  --ibs-related FILE  The input file containing related individuals according
                      to IBS value.

Options:
  --no-status         The input file doesn't have a 'status' column.

Output File:
  --out FILE          The prefix of the output files. [default: ibs_merged]

At the end of the analysis, two files are created. The file *.chosen_related_individuals" contains a list of randomly selected samples according to their relatedness (to keep only on sample for a group of related samples). The file *.discarded_related_individuals contains a list of sample to exclude to only keep unrelated samples in a dataset.

The Algorithm¶

For more information about the actual algorithms and source codes, refer to the following page.

Table Of Contents

Previous topic

Next topic

This Page