Mercurial > repos > artbio > cnv_facets
changeset 1:8e1fb7253c1d draft
planemo upload for repository https://github.com/ARTbio/tools-artbio/tree/main/tools/facets commit 613efcbc9de2f712eb2ad9540e3e63f7b968640c
| author | artbio |
|---|---|
| date | Fri, 03 Oct 2025 00:04:22 +0000 |
| parents | 3a185b812f46 |
| children | 66a56502199d |
| files | macros.xml snp_pileup_for_facets_wrapper.sh test-data/expected_pileup.csv.gz |
| diffstat | 3 files changed, 14 insertions(+), 7 deletions(-) [+] |
line wrap: on
line diff
--- a/macros.xml Wed Oct 01 01:39:32 2025 +0000 +++ b/macros.xml Fri Oct 03 00:04:22 2025 +0000 @@ -1,7 +1,7 @@ <macros> <token name="@pipefail@"><![CDATA[set -o | grep -q pipefail && set -o pipefail;]]></token> <token name="@TOOL_VERSION@">0.6.2</token> - <token name="@VERSION_SUFFIX@">0</token> + <token name="@VERSION_SUFFIX@">1</token> <token name="@PROFILE@">23.0</token> <xml name="requirements"> <requirements> @@ -10,6 +10,7 @@ <requirement type="package" version="1.22.1">samtools</requirement> <requirement type="package" version="1.22.1">htslib</requirement> <requirement type="package" version="20250822">parallel</requirement> + <requirement type="package" version="9.5">coreutils</requirement> </requirements> </xml> <xml name="citations"> <citations>
--- a/snp_pileup_for_facets_wrapper.sh Wed Oct 01 01:39:32 2025 +0000 +++ b/snp_pileup_for_facets_wrapper.sh Fri Oct 03 00:04:22 2025 +0000 @@ -113,7 +113,7 @@ "${SAMTOOLS_EXE}" index "${temp_tbam}" echo "Running snp-pileup on chromosome ${chrom}..." - "${SNP_PILEUP_EXE}" -q "${mapq}" -Q "${baseq}" ${count_orphans} "${temp_vcf}" "${temp_output}" "${temp_nbam}" "${temp_tbam}" + "${SNP_PILEUP_EXE}" --pseudo-snps=300 -q "${mapq}" -Q "${baseq}" ${count_orphans} "${temp_vcf}" "${temp_output}" "${temp_nbam}" "${temp_tbam}" } # Export all necessary variables AND the function so they are available to the sub-shells created by GNU Parallel. @@ -125,17 +125,23 @@ echo "Parallel processing finished. Concatenating results..." -# The "gather" part remains the same +# gather job outputs and sort chromosome remains the same FIRST_FILE=$(ls -1v "${TMPDIR}"/*.csv 2>/dev/null | head -n 1) if [ -z "${FIRST_FILE}" ]; then echo "Error: No pileup files were generated." >&2 exit 1 fi -(head -n 1 "${FIRST_FILE}" && \ - tail -n +2 -q "${TMPDIR}"/*.csv) | \ -bgzip > "${output_pileup}" +# Use command grouping { ...; } to pipe the combined output of head and tail. +# This entire pipeline writes the final, sorted, compressed file. +{ + # 1. Print the header once. + head -n 1 "${FIRST_FILE}"; -echo "Concatenation and compression complete." + # 2. Concatenate all files (skipping their headers). + tail -q -n +2 "${TMPDIR}"/*.csv; + +} | sort -t, -k1,1V -k2,2n | bgzip > "${output_pileup}" +echo "Concatenation, sorting and compression complete." echo "Final output is in ${output_pileup}" echo "Script finished successfully. The temporary directory will be removed by the trap."