RNAseq_Variant_Calling/tasks/Hardfiltration.wdl

task Hardfiltration {
	
	File vcf_file
	File vcf_index
	File ref_dir
	String fasta
	String sample
	String docker
	String disk_size
	String cluster_config
	

	command <<<
		set -o pipefail
		set -e
		java -Dsamjdk.use_async_io_read_samtools=false -Dsamjdk.use_async_io_write_samtools=true -Dsamjdk.use_async_io_write_tribble=false -Dsamjdk.compression_level=2 -Xmx32G -jar /gatk/gatk-package-4.1.0.0-local.jar VariantFiltration -V ${vcf_file} -O ${sample}_hc_filtered.vcf -R ${ref_dir}/${fasta} --cluster-window-size 35 --cluster-size 3 --filter-name FS --filter-expression "FS > 30.0" --filter-name QD --filter-expression "QD < 2.0" --filter-name LowDepth --filter-expression "DP < 5"
		cat ${sample}_hc_filtered.vcf |grep "#" > ${sample}_hc_filtered.vcf.header.tmp
		cat ${sample}_hc_filtered.vcf |grep PASS > ${sample}_hc_filtered.vcf.tmp
		cat ${sample}_hc_filtered.vcf.header.tmp ${sample}_hc_filtered.vcf.tmp > ${sample}_hc_PASS.vcf
		>>>

		runtime {
		docker:docker
    	cluster:cluster_config
    	systemDisk: "cloud_ssd 40"
    	dataDisk: "cloud_ssd " + disk_size + " /cromwell_root/"
	}

	output {
		File raw_vcf = "${sample}_hc_filtered.vcf"
		File pass_vcf = "${sample}_hc_PASS.vcf"
	}
}