ddbj/imputation-server-wf

The NBDC-DDBJ imputation server workflows

This repository contains the following workflows:

• Genotype imputation workflow
• Imputation reference panel construction workflow
• File format conversion workflow (PLINK to VCF)
• File format conversion workflow (binary-PLINK to VCF)

Genotype imputation workflow

Inputs of this workflow

The workflow for the genotype imputation takes two genotype datasets as inputs. The first is an experimentally observed genotype dataset (referred to as target dataset), and the second is a reference panel genotype dataset. Our workflow assumes that both genotype datasets are stored in the variant call format (VCF). We recommend users to apply quality control (QC) steps, such as missing call rate and Hardy-Weinberg equilibrium filters, to the target dataset before using this workflow. Regarding non-pseudoautosomal region on the X chromosome, this workflow assumes that male haploid genotypes are coded as “homozygous diploid”, and that the male “homozygous diploid” and female heterozygous diploid genotypes are recoded in a single unphased VCF file, according to an existing imputation workflow (https://www.protocols.io/view/genotype-imputation-workflow-v3-0-e6nvw78dlmkj/v2).

There are 2 mandatory input files.

Input name	Content
gt	Target genotype dataset (a VCF file)
region_list	Reference panel dataset (a config file that defines chunks and specifies paths to reference panel dataset (VCF and BREF3 files)

There are 2 optional input parameters.

Input name	Content
gp	Whether the posterior probability of possible genotypes is included in the output file (default is false)
nthreads	Number of threads (default value of 16)

Steps of this workflow

Step name	Content
conform-gt	The first step of the workflow detects polymorphic markers shared between the target and reference datasets using the conform-gt program (version 24May16)
beagle-imputation	The second step performs pre-phasing and genotype imputation using the beagle program version 5.2 (21Apr21.304)
beagle-vcf-index	The third step calculates the index for the imputed genotype files using bcftools (version 1.9)

The workflow assumes that pre-phasing and genotype imputation were performed using the same reference panel dataset. These three steps can be executed in parallel by splitting genomic regions into chunks. The definition of chunks is configurable by editing a text file, whereas default configurations define a chunk for each whole chromosome.

Outputs of this workflow

The output of the workflow includes the VCF files and their index files (TBI format). The VCF file contains the expected number of non-reference alleles (referred to as allele dosage) in the DS tag. Estimated allele frequencies and imputation qualities are recorded in the INFO column of the VCF files.

Output name	Content
conform-gt-vcf	Results of the step 1 (a VCF file per chunk)
conform-gt-log	Log of the step 1 (a text file per chunk)
beagle-vcf	Results of the step 2 (a VCF file per chunk)
beagle-log	Log of the step 2 (a text file per chunk)
beagle-tbi	Results of the step 3 (a TBI index file per chunk)

Imputation reference panel construction workflow

This workflow constructs an imputation reference panel (phased VCF file) from an aggregate (unphased) VCF file.

The workflow applies the following variant filtering:

• FILTER (column in the VCF) = PASS
• bi-allelic sites (remove multi-allelic sites)
• GT missingness rate (e.g., <5%)
• HWE exact test p value (e.g, > 1e-10)
• minor allele count (MAC) (e.g, ≥ 2)

After variant filtering, phasing was performed using Beagle 5.2, followed by the convertion of the phased VCF file to a BREF3 file.

Inputs of this workflow

There are 7 mandatory input files.

Input name	Content
input_vcf	Input aggregate (unphased) VCF file
samples_file	File of samples to include
gt_missingness_rate	Cut-off of GT missingness rate
hwe	Cut-off of HWE exact test p value
mac	Cut-off of minor allele count
map	Genetic map file (download from here)
region	Chromosomal region

There are 1 optional input parameter.

Input name	Content
beagle_nthreads	Number of threads for phasing (default value of 64)

Steps of this workflow

Step name	Content
stats_input	Calculate stats of the input VCF file (bcftools v1.18)
extract_samples	Extract genotype data of target samples (bcftools v1.18)
stats_extract_samples	Calculate stats of the VCF file after extracting target samples (bcftools v1.18)
remove_tags	Remove INFO tags except for INFO/AN and INFO/AC (bcftools v1.18)
fill_tags	Calculate INFO/AN, INFO/AC, INFO/HWE, INFO/ExcHet, INFO/AF, INFO/MAF, and INFO/MAC (bcftools v1.18)
filter_PASS	Exclude variants where FILTER (column in the VCF) is not 'PASS' (bcftools v1.18)
stats_filter_PASS	Calculate stats of the VCF file after excluding variants where FILTER (column in the VCF) is not 'PASS' (bcftools v1.18)
exclude_multiallelic_sites	Exclude multi-allelic variants (bcftools v1.18)
stats_exclude_multiallelic_sites	Calculate stats of the VCF file after excluding multi-allelic variants (bcftools v1.18)
filter_GT_missingness_rate	Exclude variants with low GT missingness rate (bcftools v1.18)
stats_filter_GT_missingness_rate	Calculate stats of the VCF file after excluding variants with low GT missingness rate (bcftools v1.18)
filter_HWE	Exclude variants with low HWE exact P-value (bcftools v1.18)
stats_filter_HWE	Calculate stats of the VCF file after excluding variants with low HWE exact P-value (bcftools v1.18)
filter_MAC	Exclude variants with low MAC (bcftools v1.18)
stats_filter_MAC	Calculate stats of the VCF file after excluding variants with low MAC (bcftools v1.18)
beagle_phasing	Haplotype phasing using Beagle version 5.2 (21Apr21.304)
beagle_bref3	Convert the phased VCF to BREF3 using bref3 (version 28Jun21.220)

Outputs of this workflow

The output of the workflow includes the phased VCF and BREF3 files, which can be used in the genotype imputation workflow as described above.

Output name	Content
output_vcf	A phased VCF file
beagle_phasing_log	A log file of haplotype phasing
output_bref3	A phased BREF3 file
stats_00	A file containing stats of the input VCF file
stats_01	A file containing stats of the VCF file after extracting target samples
stats_02	A file containing stats of the VCF file after excluding variants where FILTER (column in the VCF) is not 'PASS'
stats_03	A file containing stats of the VCF file after excluding multi-allelic variants
stats_04	A file containing stats of the VCF file after excluding variants with low GT missingness rate
stats_05	A file containing stats of the VCF file after excluding variants with low HWE exact P-value
stats_06	A file containing stats of the VCF file after excluding variants with low MAC

File format conversion workflow (PLINK to VCF)

Inputs of this workflow

This workflow converts PLINK file to VCF file.

There are 2 mandatory inputs.

Input name	Content
in_ped	Target PLINK (PED) file path
out_name	Output VCF file name

File format conversion workflow (binary-PLINK to VCF)

Inputs of this workflow

This workflow converts binary-PLINK file to VCF file.

There are 2 mandatory inputs.

Input name	Content
in_bed	Target binary-PLINK (BED) file path
out_name	Output VCF file name