bash script for imputing a genomic interval. Used with the SGE job array script.

impute2_job.sh

#!/bin/bash

CHR=$1
INTERVAL_START=`printf "%.0f" $2`
INTERVAL_END=`printf "%.0f" $3`
INTERVAL=`printf "%03d" $4`

# directories
NAME="impute_test_data_hg19"
OUTPUT_DIR="/mnt/iusers01/jw01/mdeasjdb/scratch/impute_practice/"
PHASE_DATA_DIR="phasing/output/"
REF_DIR="/mnt/iusers01/jw01/jw01-shared-resources/impute2/ref_panel/ALL_1000G_phase1integrated_v3_impute_macGT1/"

# executable
IMPUTE2_EXEC=/mnt/iusers01/jw01/jw01-shared-resources/impute2/bin/impute2_2.3.0

# parameters
K=80
ITER=30
BURNIN=10
NE=20000

# Genetic map.
MAP_FILE=${REF_DIR}genetic_map_chr${CHR}_combined_b37.txt

# Panel 0 phased reference haplotypes.
REF_HAP_FILE=${REF_DIR}ALL_1000G_phase1integrated_v3_chr${CHR}_impute_macGT1.hap.gz
REF_LEGEND_FILE=${REF_DIR}ALL_1000G_phase1integrated_v3_chr${CHR}_impute_macGT1.legend.gz

# Panel 2 phased inference genotypes.
STUDY_HAP_FILE=${PHASE_DATA_DIR}${NAME}_chr${CHR}.haps.gz
STUDY_SAMPLE_FILE=${PHASE_DATA_DIR}${NAME}_chr${CHR}.sample

# main output file
OUTPUT_FILE=${OUTPUT_DIR}${NAME}_chr${CHR}.int${INTERVAL}.pos${INTERVAL_START}-${INTERVAL_END}.impute2

# Impute interval
$IMPUTE2_EXEC \
    -use_prephased_g \
    -known_haps_g $STUDY_HAP_FILE \
    -h $REF_HAP_FILE \
    -l $REF_LEGEND_FILE \
    -m $MAP_FILE \
    -int $INTERVAL_START $INTERVAL_END \
    -Ne $NE \
    -k $K \
    -iter $ITER \
    -burnin $BURNIN \
    -filt_rules_l 'eur.maf<0.01' \
    -o $OUTPUT_FILE \
    -o_gz