As a complementary pipeline for rare variant analysis on GWAS chips, EGRVA is straightforward and cost-efficient.

Table 1. List of pre-requisite software and the available information.

Make sure the pre-requisite software listed in Table 1 has been installed.

1. Pre-QC executed by PLINK.

   $Plink --bfile file1 --keep-allele-order --hwe 0.00001 --geno 0.05 --maf 0.005 --max-maf 0.01 --mind 0.05 --make-bed --out file2
   

2. Use LiftOver to convert the genome coordinates into hg19.

   python liftOverPlink.py -m file.map -p file.ped -o plink_ped -c hg18ToHg19.over.chain.gz -e liftOver

   NOTE:

   "liftOverPlink.py" and "hg18ToHg19.over.chain.gz" can be downloaded from https://github.com/ruijiali/EGRVA. If your dataset build is hg19, skip this step.

3. Use SHAPEIT2 as the phasing tool.

4. Check the PLINK bim file against the HRC reference SNP list with the script .

   $PLINK --bfile file --freq
   

   $perl HRC-1000G-check-bim.pl -b file.bim -f plink.frq -r HRC.r1-1.GRCh37.wgs.mac5.sites.vcf -h

   NOTE:

   "HRC-1000G-check-bim.pl" can be downloaded from https://github.com/ruijiali/EGRVA.

Sign up the Michigan Imputation service (https://imputationserver.sph.umich.edu/index.html) and upload vcf files separated by chromosomes. Generally, the filling operation time will not be too long.

NOTE:

We recommend HRC r1.1 2016 for Reference Panel, 0.3 for rsq Filter, and Population according to your samples.

1. Filter SNPs with R2 < 0.8.

   for i in {1..22}; do bcftools view -i 'R2>.8' -Oz chr${i}.dose.vcf.gz > file1; done
   

2. Changing the ID to '%CHROM:%POS:%REF:%ALT'.

   for i in {1..22}; do bcftools norm -Ou -m+any file1| bcftools annotate --output-type b --output file2  -I '%CHROM:%POS:%REF:%ALT'; done
   

3. Cut

   do bcftools view file2| cut -f3  | awk '{a[$0]++; if(a[$0]==2) print; if (a[$0]>=2) print}'; 
   

4. Convert bcf format to plink format.

   $PLINK --bcf file2--keep-allele-order --allow-extra-chr 0 --const-fid --split-x b37 no-fail --vcf-idspace-to _ --make-bed --out file4 
   

5. Check for duplicates and deletions.

   Run "duplicate.R" in R software, R reports "No duplicate SNPs".

   NOTE:

   "duplicate.R" can be downloaded from https://github.com/ruijiali/EGRVA.

6. Merge chromosomes.

   $PLINK --bfile file4 --merge-list mergelist.txt --biallelic-only --make-bed --out file5
   

7. Post-QC executed by PLINK.

   $PLINK --bfile file5 --keep-allele-order --hwe 0.00001 --geno 0.05 --maf 0.005 --max-maf 0.01 --make-bed --mind 0.05 --noweb --out file6
   

8. Remove multiallelic SNPs.

   $Plink --bfile file6 --snps-only --make-bed --out file7
   

1. Convert plink to VCF4 file.

   $PLINK --bfile file7 --recode vcf-iid --out file8
   

2. ANNOVAR annotation

   perl convert2annovar.pl -format vcf4old file8 > file9;
   perl file9 humandb/ -buildver hg19 -out file10 -remove -protocol related database -operation g,f
   

1. Count allele frequency seperately by case and control.

   $Plink --bfile file10 --keep case_samole_file -extract exonic_file  --recode vcf-iid --out  file11_case
   $Plink --bfile file10 --keep control_samole_file -extract exonic_file  --recode vcf-iid --out  file11_control
   

2. Combine to get the file"12.csv"

3. Run "mirage.R" in R software.

   NOTE:

   "mirage.R" can be downloaded from https://github.com/ruijiali/EGRVA. MIRAGE is available at https://xinhe-lab.github.io/mirage.

Table 2. List of bioinformatics analysis tools and the available information.