mendelian_concordance/tasks/merge.wdl

task merge {
	Array[File] family_vcf_gz
	Array[File] family_vcf_idx
	String test_name
	String sample
	String docker
	String cluster_config
	String disk_size
	
	command <<<

		rtg vcfmerge --force-merge-all -o ${sample}.merged.vcf.gz ${sep=" " family_vcf_gz}
		rtg vcffilter -i ${sample}.merged.vcf.gz -o ${sample}.snv.merged.vcf.gz --snps-only --all-samples
		rtg vcffilter -i ${sample}.merged.vcf.gz -o ${sample}.indel.merged.vcf.gz --non-snps-only --all-samples

		zcat ${sample}.indel.merged.vcf.gz | grep '#CHROM' | cut -f10-12 > name

		for i in {10..12}; do zcat ${sample}.snv.merged.vcf.gz | grep -v '#' | cut -f$i | cut -d ':' -f2-4 | grep -v '\.'| sort | uniq -c | awk '{print $1, substr($1,0,7)}' | sed 's/\s\+/\t/g' | cut -f1 > $i.snv.txt; done

		paste *.snv.txt | cat name - > snv.txt

		for i in {10..12}; do zcat ${sample}.indel.merged.vcf.gz | grep -v '#' | cut -f$i | cut -d ':' -f2-4 | grep -v '\.'| sort | uniq -c | awk '{print $1, substr($1,0,7)}' | sed 's/\s\+/\t/g' | cut -f1 > $i.indel.txt; done

		paste *.indel.txt | cat name - > index.txt

		echo 'type' > column
		echo '0,0,0' >> column
		echo '0,0,1' >> column
		echo '0,1,0' >> column
		echo '0,1,1' >> column
		echo '1,0,0' >> column
		echo '1,0,1' >> column
		echo '1,1,0' >> column
		echo '1,1,1' >> column

		paste column snv.txt > ${test_name}.snv.txt

		paste column indel.txt > ${test_name}.indel.txt


	>>>

	runtime {
		docker:docker
		cluster: cluster_config
		systemDisk: "cloud_ssd 40"
		dataDisk: "cloud_ssd " + disk_size + " /cromwell_root/"
	}
	output {
		File merged_vcf = "${sample}.merged.vcf.gz"
		File merged_vcf_idx = "${sample}.merged.vcf.gz.tbi"
		File merged_snv = "${sample}.snv.merged.vcf.gz"
		File merged_snv_idx = "${sample}.snv.merged.vcf.gz.tbi"
		File merged_indel = "${sample}.indel.merged.vcf.gz"
		File merged_indel_idx = "${sample}.indel.merged.vcf.gz.tbi"
		File snv = "${test_name}.snv.txt"
		File indel = "${test_name}.indel.txt"
	}
}