calc_pca2_projection.sh

#!/bin/bash
set -e
source constants_.sh
source parse_args.sh "$@"

# Parse input
target_dataset="${datasets_path}${target}/${imp}/"
# ref_dataset="${datasets_path}${target%%_*}/${imp}/"
ref_dataset=${datasets_path}/${ref_dataset}/
if [[ -z ${maf} ]]; then maf=0.05; fi
if [[ -z ${geno} ]]; then geno=0.1; fi
if [[ -z ${imp} ]]; then imp="original"; fi
if [[ -z ${memory} ]]; then memory=500000; fi
if [[ -z ${threads} ]]; then threads=80; fi
if [[ -z ${stage} ]]; then stage=4; fi

# Start pipeline

if [[ ${stage} -le 1 ]]; then
 	echo '### perform ref prunning ###'
 	plink \
 		--bfile ${ref_dataset}ds.QC \
 		--out ${ref_dataset}ds \
 		--memory ${memory} \
 		--threads ${threads} \
 		--indep-pairwise 200 50 0.25
fi

if [[ ${stage} -le 2 ]]; then
	echo '### calc ref pca'
	plink2 \
		--bfile ${ref_dataset}ds.QC \
		--out ${ref_dataset}ds.ref \
		--memory ${memory} \
		--threads ${threads} \
		--extract ${ref_dataset}ds.prune.in \
		--freq counts \
		--make-rel \
		--pca approx allele-wts 6
fi
#if [[ ${stage} -le 2 ]]; then
#	echo '### QC ###'
#	plink \
#		--bfile ${target_dataset}ds \
#		--out ${target_dataset}ds.QC \
#		--memory ${memory} \
#		--threads ${threads} \
#		--maf ${maf} \
#		--geno ${geno} \
#		--hwe 1e-6 \
#		--make-bed
#fi


# if [[ ${stage} -le 3 ]]; then
# 	echo '### perform prunning ###'
# 	plink \
# 		--bfile ${target_dataset}ds.QC \
# 		--out ${target_dataset}ds \
# 		--memory ${memory} \
# 		--threads ${threads} \
# 		--indep-pairwise 200 50 0.25
# fi
# if [[ ${stage} -le 4 ]]; then
# 	echo  "calc target pca"
# 	plink2 \
# 		--bfile ${target_dataset}ds.QC \
# 		--out ${target_dataset}ds.pca \
# 		--memory ${memory} \
# 		--threads ${threads} \
# 		--extract ${ref_dataset}ds.prune.in \
# 		--geno 0.1 \
# 		--mind 0.1 \
# 		--read-freq ${ref_dataset}ds.ref.acount \
# 		--score ${ref_dataset}ds.ref.eigenvec.allele 2 5 header-read no-mean-imputation variance-standardize \
# 		--score-col-nums 6-11
# fi
if [[ ${stage} -le 4 ]]; then
	echo  "calc target pca"
	plink2 \
		--bfile ${target_dataset}ds.QC \
		--out ${target_dataset}ds.pca \
		--memory ${memory} \
		--threads ${threads} \
		--extract ${ref_dataset}ds.prune.in \
		--geno 0.1 \
		--mind 0.1 \
		--read-freq ${ref_dataset}ds.ref.acount \
		--score ${ref_dataset}ds.ref.eigenvec.allele 2 5 header-read no-mean-imputation  \
		--score-col-nums 6-11
fi