Mercurial > repos > iuc > sra_tools

changeset 19:248f85ff0733 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/sra-tools commit 1e2492fa4557ddc6ffefc16b249ed9ef0207d5e1"
author iuc
date Tue, 23 Jun 2020 17:18:45 -0400
parents 7068f48d0ef9
children 964579f93c54
files fasterq_dump.xml fastq_dump.xml sam_dump.xml sra_macros.xml test-data/SRR11953971_1.fastq.gz test-data/SRR11953971_2.fastq.gz test-data/sra_manifest.tabular
diffstat 7 files changed, 34 insertions(+), 4 deletions(-) [+]
line wrap: on
line diff
--- a/fasterq_dump.xml	Mon Jun 08 05:49:21 2020 -0400
+++ b/fasterq_dump.xml	Tue Jun 23 17:18:45 2020 -0400
@@ -1,4 +1,4 @@
-<tool id="fasterq_dump" name="Faster Download and Extract Reads in FASTQ" version="@VERSION@+galaxy0" profile="18.01">
+<tool id="fasterq_dump" name="Faster Download and Extract Reads in FASTQ" version="@VERSION@+galaxy1" profile="18.01">
     <description>format from NCBI SRA</description>
     <macros>
         <import>sra_macros.xml</import>
@@ -169,6 +169,28 @@
                 <element name="SRR002702" file="SRR002702_2.fastq.gz" ftype="fastqsanger.gz" decompress="True"/>
             </output_collection>
         </test>
+        <test>
+            <param name="input_select" value="file_list"/>
+            <param name="file_list" value="sra_manifest.tabular" ftype="sra_manifest.tabular"/>
+            <output_collection name="list_paired" type="list:paired">
+                <element name="SRR11857975">
+                    <element name="forward" file="SRR11857975_1.fastq.gz" decompress="True">
+                    </element>
+                    <element name="reverse" file="SRR11857975_2.fastq.gz" decompress="True">
+                    </element>
+                </element>
+            </output_collection>
+        </test>
+        <test>
+            <param name="input_select" value="file_list"/>
+            <param name="file_list" value="sra_manifest.tabular" ftype="sra_manifest.tabular"/>
+            <output_collection name="list_paired" type="list:paired">
+                <element name="SRR11953971">
+                    <element name="forward" file="SRR11953971_1.fastq.gz" decompress="True"/>
+                    <element name="reverse" file="SRR11953971_2.fastq.gz" decompress="True"/>
+                </element>
+            </output_collection>
+        </test>
     </tests>
     <help><![CDATA[
 **What it does?**
--- a/fastq_dump.xml	Mon Jun 08 05:49:21 2020 -0400
+++ b/fastq_dump.xml	Tue Jun 23 17:18:45 2020 -0400
@@ -1,4 +1,4 @@
-<tool id="fastq_dump" name="Download and Extract Reads in FASTA/Q" version="@VERSION@+galaxy0" profile="18.01">
+<tool id="fastq_dump" name="Download and Extract Reads in FASTA/Q" version="@VERSION@+galaxy1" profile="18.01">
     <description>format from NCBI SRA</description>
     <macros>
         <import>sra_macros.xml</import>
--- a/sam_dump.xml	Mon Jun 08 05:49:21 2020 -0400
+++ b/sam_dump.xml	Tue Jun 23 17:18:45 2020 -0400
@@ -1,4 +1,4 @@
-<tool id="sam_dump" name="Download and Extract Reads in BAM" version="@VERSION@+galaxy0" profile="18.01">
+<tool id="sam_dump" name="Download and Extract Reads in BAM" version="@VERSION@+galaxy1" profile="18.01">
     <description>format from NCBI SRA</description>
     <macros>
         <import>sra_macros.xml</import>
--- a/sra_macros.xml	Mon Jun 08 05:49:21 2020 -0400
+++ b/sra_macros.xml	Tue Jun 23 17:18:45 2020 -0400
@@ -19,7 +19,13 @@
     ]]></token>
     <token name="@SET_ACCESSIONS@"><![CDATA[
         #if $input.input_select=="file_list":
-            for acc in `@ACCESSIONS_FROM_FILE@ '$input.file_list'` ;
+            #if $input.file_list.is_of_type('sra_manifest.tabular'):
+                #set $column = $input.file_list.unsanitized.metadata.column_names.index('Run') + 1
+                cut -f $column '$input.file_list'| tail -n +2 > "manifest" &&
+            #else
+                ln -s '$input.file_list' > manifest &&
+            #end if
+            for acc in `@ACCESSIONS_FROM_FILE@ manifest` ;
             do (
         #elif $input.input_select=="accession_number":
             acc='${input.accession}' && [ ""\$acc" =~ ^[E|S|D]RR[0-9]{1,}$" ] && (
Binary file test-data/SRR11953971_1.fastq.gz has changed
Binary file test-data/SRR11953971_2.fastq.gz has changed
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/test-data/sra_manifest.tabular	Tue Jun 23 17:18:45 2020 -0400
@@ -0,0 +1,2 @@
+Run	Assay Type	BioProject	BioSample	Center Name	Consent	DATASTORE filetype	DATASTORE provider	DATASTORE region	Experiment	Instrument	LibraryLayout	LibrarySelection	LibrarySource	Organism	Platform	Sample Name	SRA Study	Bytes	ReleaseDate	Library Name	Collection_Date	isolate	AvgSpotLen	Bases	BioSampleModel	Host	geo_loc_name_country	geo_loc_name_country_continent	geo_loc_name	host_disease	collected_by	isolation_source	lat_lon	host_sex	passage_history	strain	host_age	host_subject_id	ENA-FIRST-PUBLIC (run)	ENA-FIRST-PUBLIC	ENA-LAST-UPDATE (run)	ENA-LAST-UPDATE	geographic_location_(country_and/or_sea)	host_health_state	collecting_institution	collector_name	geographic_location_(region_and_locality)	host_common_name	host_scientific_name	gisaid_id	isolation_source_host-associated	sample_capture_status	isolate (run)	ARTIC_barcode_identifiers	host_tissue_sampled	library_id	AssemblyName	GEO_Accession (exp)	source_name	tissue	barcode_identifiers	Cell_type	host_description	sample_type	age_event	Age	Biomaterial_provider	cell_number	cell_processing_protocol	cell_quality	disease	disease_stage	ETHNICITY	health_state	sex	single_cell	synthetic	tissue_processing	cell_subsets	patient_group	disease_state	subject_status	tissue/cell_type	sample_origin	organ	host_disease_stage	stock_production_date	culture_collection	COUNTRY	Extraction-method	host_disease_outcome	identification_method	Laboratory_Host	link_addit_analys	ref_biomaterial	datatype (run)
+SRR11953971	RNA-Seq	PRJNA622837	SAMN14938999	BROAD INSTITUTE OF HARVARD AND MIT	public	bam,sra	gs,ncbi,s3	gs.US,ncbi.public,s3.us-east-1	SRX8498148	Illumina NovaSeq 6000	PAIRED	cDNA	VIRAL RNA	Severe acute respiratory syndrome coronavirus 2	ILLUMINA	MA_MGH_00524	SRP266465	207470	2020-06-08T00:00:00Z	SAMN14938999_ERCC-00040_RandomHexamer_NexteraXT	2020-03-17	not applicable	202	415514	Pathogen.cl	Homo sapiens	USA	North America	USA: Massachusetts	COVID-19	Massachusetts General Hospital	nasopharyngeal swab	missing		not applicable	not applicable		MA_MGH_00524