ancestry

#!/bin/bash

# Step 1: Input Sanitation

# 1.1 
REFPANEL=$1



# 1.2
# vcf/ped in current dir -> me.ped
input="$(ls | egrep -e ".vcf|.vcf.gz|.ped")"

echo $input
if [[ "$input" == *ped ]] 
then
    filename="$(echo $input | cut -f1 -d.)" 
    ./plink --file $filename --recode --out me > run.log 2>error.log
elif [ "$input" == *vcf ] || [ "$input" == *gz ] ;
then
    ./plink --vcf $input --maf 0.05 --recode --out me > run.log 2>>error.log
else
    echo No vcf/ped file found in the current directory
    exit 1
fi



# 1.3
# me.ped -> me_binary.bed
./plink --file me --make-bed --out me_binary >> run.log 2>>error.log



# Step 2: Merging Efficiently
# me_binary.bed $REFPANEL.bed -> merge.bed

# 2.1
awk '{print $2;}' me_binary.bim | sort > me_binary.snps
awk '{print $2;}' reference_panel/$REFPANEL".bim" | sort > ref.snps

# 2.2
comm -12 me_binary.snps ref.snps > merging.snps

# 2.3
./plink --bfile reference_panel/$REFPANEL --bmerge me_binary --extract merging.snps -make-bed --out merge1 >>run.log 2>>error.log

# 2.4
./plink --bfile merge1 -geno 0.9 --make-bed --out merge >>run.log 2>>error.log


# Step 3: Preparing requisite files
# merge.bed -> merge.pop -> $pops

# 3.1
awk '{print $1}' merge.fam > merge1.pop

# 3.2
awk '{ if( $1=="FAM001" ) $1=" "; print $1; }' merge1.pop > merge.pop

# 3.3
pops="$(cat merge.pop | sort | uniq | wc -l)"
let pop=pops-1


# Step 4: Admixture analysis and plotting

# 4.1
./admixture -j5 --supervised merge.bed $pop >> run.log 2>>error.log >> run.log 2>>error.log

# 4.2
output="$(ls | egrep -e "$pop.Q")"

# 4.3
Rscript plots/plot.R $output 1>>error.log

# Review results and house keeping
open ancestry.png
echo
rm -f me_binary.snps ref.snps merging.snps