example to pass bash associative arrays to functions

get-last-passing-line-in-vcf.sh

#!/bin/bash 

# Example usage: bash get-passing-line-in-vcf.sh /gscmnt/gc2758/analysis/ccdg_gatk_callsets/eocad_aa_2017.05.18/post-vqsr/12-concat-vcfs

# programs
BCFTOOLS=/gscmnt/gc2802/halllab/idas/software/local/bin/bcftools1.4
TABIX=/gscmnt/gc2802/halllab/idas/software/local/bin/tabix
SORT=/gscmnt/gc2802/halllab/idas/software/local/bin/lh3-sort

# data
B38_FAI=/gscmnt/gc2802/halllab/ccdg_resources/genomes/human/GRCh38DH/all_sequences.fa.fai

function get_chrom_lengths {
    local -A lengths
    while IFS=$'\t' read -r chrom size; do
        lengths[${chrom}]=${size}
    done < <(head -n 24 ${B38_FAI} | cut -f 1,2)
    declare -p lengths | sed -e 's/declare -A \w\+=//'
}

function count_passing_variants {
    local region=$1
    local vcf=$2
    ${BCFTOOLS} query -r ${region} -f '%POS\t%FILTER\n' ${vcf} | grep PASS | wc -l
}

function get_final_passing_variants {
    local region=$1
    local vcf=$2
    ${BCFTOOLS} query -r ${region} -f '%POS\t%FILTER\n' ${vcf} | grep PASS | tail -n 1
}

function get_last_entry {
    local chrom=$1
    local max=$2
    local vcf=$3

    local step=10000

    local start=$(( ${max} - ${step} ))
    region=${chrom}:${start}-${max}

    local count=$(count_passing_variants ${region} ${vcf})
    while [[ ${count} == 0 ]]; do
        start=$(( ${start} - ${step} ))
        region=${chrom}:${start}-${max}
        count=$(count_passing_variants ${region} ${vcf})
#        echo '============'
#        ${BCFTOOLS} query -r ${region} -f '%POS\t%FILTER\n' ${vcf}
#        echo '============'
    done
    local out=$(get_final_passing_variants ${region} ${vcf})
    printf "chrom : %s | last-position : %s\n" "${chrom}" "${out}"
}

function main {
    local vcfdir=$1

    local lengths_string=$(get_chrom_lengths)
    eval "declare -A lengths=${lengths_string}"

    local -a chroms=($(for k in "${!lengths[@]}"; do echo ${k}; done | ${SORT} -N))

    for c in ${chroms[@]}; do
        local vcf=${vcfdir}/${c}/concatenated.c${c}.vcf.gz
        get_last_entry ${c} ${lengths[${c}]} ${vcf}
    done
}

main "$@";

# declare -p NAME : display the attributes and values of each NAME
# https://stackoverflow.com/questions/10582763/how-to-return-an-array-in-bash-without-using-globals