GWASQC

Genotyping Quality Control focused on admixed populations. This pipeline is part of a LARGE-PD variant calling and QC procedure, specifically designed for LARGE-PD Phase 2 data.

Options

Genotyping QC to GWAS

optional arguments:
  -h, --help            show this help message and exit

Required arguments:
  -i INPUTFILE, --inputFile INPUTFILE
                        Input file without suffix
  -o OUTPUTNAME, --outputName OUTPUTNAME
                        Output name
  -O OUTPUTFOLDER, --outputFolder OUTPUTFOLDER
                        Output folder
  -I INFO, --info INFO  File with covar information. The pipeline requires at least three columns (ID, Sex, Phenotype)

Optional arguments:
  -e, --erase           Set to exclude all temporary files
  -p PHENONAME, --phenoName PHENONAME
                        Name of the phenotype column in the info file (default: PHENO)
  -s STEPS, --steps STEPS
                        File with the order of the steps (one step per line)
  -P POPFILE, --popFile POPFILE
                        File with two columns: Individual ID and Population ID.This approach is important
                        forrelationship control that should be done per population
  -S SAVEPERPOP, --savePerPop SAVEPERPOP
                        Save the QCed data with all pops and per population

Programs arguments:
  --plink2 PLINK2       Path to Plink 2 (default: plink2)
  --plink1 PLINK1       Path to Plink 1.9 (default: plink)
  --NAToRA NATORA       Path to NAToRA (default: NAToRA.py)
  --python PYTHON       Python > 3 with networkX instaled (default: python)
  --king KING           Path to king program (default: king)

Input files

• Input: Data with genotyped information. It can be VCF(vcf or vcf.gz), Plink Bfile (default to plink 1.90) or Plink Pfile (default to plink 2)
• Steps: List of QC steps to be performed and the parameters. If you did provide any file, the program will run the default QC with default parameters. See QC Steps Section.
• Pop file: File with two columns. The first column is the individual ID and the second is the population ID. This file is used by relationship control and to the flag -S (to be implemented)
• Info: Provide covariate information in a file with at least 3 columns and a header:

  • ID: Sample ID as in the input data. Possible header names: ID, ind, sample_external_id.

  • SEX: Sex information for the sample. Possible header names: Sex, gender.

    • We accept four different values for SEX:
      • Males:
        • 0 (0/1 codification)
        • 1 (1/2 codification, default to Plink)
        • M or Male
      • Females:
        • 1 (0/1 codification)
        • 2 (1/2 codification, default to Plink)
        • F or Female

  • STATUS: Status information for the sample. Possible header names: Pheno, status, affection_status.

    • Case:

      • 1 (0/1 codification)
      • 2 (1/2 codification, default to Plink)
      • case, pd case, affected

    • Control:

      • 0 (0/1 codification)
      • 1 (1/2 codification, default to Plink)
      • control, pd control, unaffected, non-affected, non affected, not affected

Example for info file:

ID,SEX,STATUS
Sample1,M,Case
Sample2,F,Control
Sample3,Male,Control
Sample4,Female,Affected

Our code is not case-sensitive, and it accommodates different conventions for header names. If your file has headers with names like "id" instead of "ID" our pipeline will handle it seamlessly.

QC Steps

In this section, we provide a brief overview of the quality control (QC) steps performed by the pipeline. The steps are presented in the default order.

Removal of Samples without Sex and Status (Mandatory)

• Description:

  • We prioritize data integrity by requiring sex and status information for all samples. Samples lacking this essential information are removed from the analysis.

• Implementation:

  • We remove all samples with missing data in the Sex and Status columns of the covariate info file.

Sex-Check (Optional)

• Description:

  • This step involves checking the consistency between the declared sex (provided in the info file) and the inferred sex based on the sexual chromosome information using Plink1.90.

• Implementation:

  • We use Plink1.90 with specific thresholds to determine sex discrepancies.

    • Females: F < 0.5
    • Males: F > 0.8

  • Note:

    • By default, Plink uses F < 0.2 to female. In admixed populations, females might commonly have F statistics higher than 0.2, what can cause an unnecessary sample loss.

Redundant variants (Optional)

• Description:
  • The genotyping platform used in LARGE-PD Phase 2 sometimes reports genotypes for variants on the "+" strand and other times on "-" strand. This can lead to issues, especially for variants involving A/T and C/G, potentially introducing bias into the analysis.
• Implementation:
  • We remove all variants that are A/T or C/G.

Sample missing data (Optional)

• Description:

  • High missing data in samples may indicate issues such as problems with DNA extraction, degraded DNA, or low DNA concentration, which can impact the reliability of genotyping data.

• Implementation:

  • To ensure data quality, we exclude samples with missing data exceeding a specified cutoff.
    • Default cutoff: 0.05 (5%)

Variant missing data (Optional)

• Implementation:
  • To ensure data quality, we exclude variants with missing data exceeding a specified cutoff.
    • Default cutoff: 0.05 (5%)

Duplicate variants (Optional)

• Description:

  • Certain analyses may require the absence of duplicate variants to avoid redundancy and potential biases in downstream results.

• Implementation:

  • We identify and compare duplicated variants, calculating the missing data for each duplicate. To maintain data quality, we keep the variant with the least missing data.

Heterozigosity (Optional)

• Description:

  • Deviations in heterozygosity can serve as indicators of sample contamination or inbreeding within the dataset.

• Implementation:

  • We calculate the sample heterozygosity and provide an optional step to remove samples that deviate from a specified cutoff.
    • Default cutoff: 3 (sd).

• Note:

  • While some studies remove samples based on Plink F statistics, it's important to note that this approach in admixed populations can result in unnecessary sample loss.

HWE in Controls (Optional)

• Description:

  • Deviations from Hardy Weinberg Equilibrium (HWE) can indicate issues in the genotyping process. It is a common practice to set different HWE cutoffs for cases and controls.

• Implementation:

  • Variants with p-values below the specified cutoff are removed.
    • Default cutoff: 1e-6.

HWE in Cases (Optional)

• Description:

  • Deviations from Hardy Weinberg Equilibrium (HWE) can indicate issues in the genotyping process. It is a common practice to set different HWE cutoffs for cases and controls.

• Implementation:

  • Variants with p-values below the specified cutoff are removed.
    • Default cutoff: 1e-10.

Relationship Control (Optional)

• Description:

  • The presence of related samples can cause bias in genetic analysis.

• Implementation:

  • We use NAToRA software to remove related samples. We consider related samples which kinship coefficient > cutoff
    • Default cutoff: 0.0884 (second degree).

• Note:

  • If your data have samples from different population, the best standard is to calculate the relationship by population

Example file to --steps following the default order and cutoffs:

sex-check
ATCG
mind	0.05
geno	0.05
duplicate
heterozygosity	3
HWE control	1e-6
HWE case	1e-10
relationship	0.0884

Programs used

In this pipeline we use:

• Plink1
• Plink2
• Python3
  • Libraries required:
    • NetworkX *
    • Numpy
• NAToRA (Leal et al. 2022)*

* Used on relationship control

Common Problems

• PLINK2 errors:

  • Error: Unrecognized flag ('--het').
    • PLINK2 is a tool in development, so we strongly recommend download the latest version and use it.

• NAToRA errors:

  • ImportError: No module named networkx
    • Install networkX in your python.
      • pip install networkx or python -m pip install networkx
    • If the problem persists
      • Some computer environments have different python installed. To fix this
        • type which python and copy the path and add on the --python in your command line

Acknowledgements

This work is supported by NIH Grant R01 1R01NS112499-01A1, MJFF Grant ID: 18298, ASAP-GP2 and Parkinson's Foundation

We also want to thank all users that report problems/bugs.

Contact

Developer: Thiago Peixoto Leal. PhD (PEIXOTT@ccf.org or thpeixotol@hotmail.com)