Mercurial > repos > trinity_ctat > ctat_genome_resource_libs_data_manager_3

changeset 52:0e3e4f24e300 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Adding ctat_cravat_tissues table.
author trinity_ctat
date Tue, 06 Nov 2018 11:53:38 -0500
parents 0a4bec29a609
children 9fdaeb356e22
files data_manager/add_ctat_resource_lib.py data_manager/add_ctat_resource_lib.xml data_manager_conf.xml tool-data/cravat_tissues.loc.sample tool-data/ctat_cravat_tissues.loc.sample tool_data_table_conf.xml.sample
diffstat 6 files changed, 68 insertions(+), 29 deletions(-) [+]
line wrap: on
line diff
--- a/data_manager/add_ctat_resource_lib.py	Fri Nov 02 14:17:08 2018 -0400
+++ b/data_manager/add_ctat_resource_lib.py	Tue Nov 06 11:53:38 2018 -0500
@@ -1324,6 +1324,12 @@
              'Normally they are assumed to reside in the build directory, ' + \
              'but if that directory has not been created yet when this program ' + \
              'is called, you can specify the full path to the directory where they reside.')
+    parser.add_argument('-t', '--cravat_tissues_filepath', 
+        default='', \
+        help='Specify a non-default location where the Cosmic files reside. ' + \
+             'Normally they are assumed to reside in the build directory, ' + \
+             'but if that directory has not been created yet when this program ' + \
+             'is called, you can specify the full path to the directory where they reside.')
     # Method 1) arguments - Download and Build. 
     # - One can optionally utilize --build_location argument with this group of arguments.
     download_and_build_args = parser.add_argument_group('Download and Build arguments')
@@ -1546,6 +1552,17 @@
     data_manager_dict['data_tables']['ctat_genome_resource_libs'] = []
     data_table_entry = dict(value=unique_id, name=display_name, path=genome_build_directory)
     data_manager_dict['data_tables']['ctat_genome_resource_libs'].append(data_table_entry)
+    
+    # Create the data table for the cravat_tissues, if the file is given:
+    if (args.cravat_tissues_filepath is not None) and (args.cravat_tissues_filepath != ""):
+        data_manager_dict['data_tables']['ctat_cravat_tissues'] = []
+        cravat_file = open(args.cravat_tissues_filepath, 'r')
+        for line in cravat_file:
+            if line[0] != '#':
+                # The line is not a comment, so parse it.
+                items = [item.strip() for item in line.split("\t")]
+                data_table_entry = dict(value=items[0], name=items[1], code=items[2], date=items[3])
+                data_manager_dict['data_tables']['ctat_cravat_tissues'].append(data_table_entry)
 
     # Temporarily the output file's dictionary is written for debugging:
     print "The dictionary for the output file is:\n\t{:s}".format(str(data_manager_dict))
--- a/data_manager/add_ctat_resource_lib.xml	Fri Nov 02 14:17:08 2018 -0400
+++ b/data_manager/add_ctat_resource_lib.xml	Tue Nov 06 11:53:38 2018 -0500
@@ -14,7 +14,8 @@
         <![CDATA[
           python $__tool_directory__/add_ctat_resource_lib.py 
             --output_filename="${out_file}" 
-            --display_name="${display_name}"            
+            --display_name="${display_name}"
+            --cravat_tissues_filepath="${__tool_directory__}/../tool-data/ctat_cravat_tissues.loc.sample"           
             #if str( $genome_resource_library.build_type ) == "download_and_build":
               --download_url="${genome_resource_library.download_url}" 
               --download_location="${genome_resource_library.download_destination}"
@@ -188,7 +189,8 @@
 
         Neither the "source_data" nor the "plug-n-play" versions have Mutation Resources included. 
         Those must be downloaded separately and integrated into the Library. If you are going to be using the 
-        ctat_mutations tool, check the Download Mutation Library check box.
+        ctat_mutations tool, check the Download Mutation Library check box. Whether or not you check this box, the
+        ctat_cravat_tissues table needed by the ctat_mutations tool will be created.
 
         In order to integrate the Mutation Resources into a CTAT Genome Resource Library, you must have previously downloaded
         COSMIC resources (See Step 2 from https://github.com/NCIP/ctat-mutations/tree/master/mutation_lib_prep )
--- a/data_manager_conf.xml	Fri Nov 02 14:17:08 2018 -0400
+++ b/data_manager_conf.xml	Tue Nov 06 11:53:38 2018 -0500
@@ -34,5 +34,21 @@
                 <!--</column> -->
             </output>
         </data_table>
+        <data_table name="ctat_cravat_tissues">
+            <output>
+                <column name="value" />
+                    <!-- value is used to uniquely identify this entry in the table.
+                    -->
+                <column name="name" />
+                    <!-- name is used as the selector in the pull down lists for items in this table.
+                    -->
+                <column name="code" />
+                    <!-- .
+                    -->
+                <column name="date" />
+                    <!-- .
+                    -->
+            </output>
+        </data_table>
     </data_manager>
 </data_managers>
--- a/tool-data/cravat_tissues.loc.sample	Fri Nov 02 14:17:08 2018 -0400
+++ /dev/null	Thu Jan 01 00:00:00 1970 +0000
@@ -1,27 +0,0 @@
-Bladder   Bladder Urothelial Carcinoma  BLCA (TCGA)   Jun 2013
-Blood-Lymphocyte  Chronic Lymphocytic Leukemia  CLL (ICGC)  Mar 2013
-Blood-Myeloid   Acute Myeloid Leukemia  LAML (TCGA)   Jun 2013
-Brain-Cerebellum  Medulloblastoma   MB (mixed source)   Dec 2010
-Brain-Glioblastoma-Multiforme   Glioblastoma Multiforme   GBM (TCGA)  Jun 2013
-Brain-Lower-Grade-Glioma  Brain Lower Grade Glioma  LGG (TCGA)  Jun 2013
-Breast  Breast Invasive Carcinoma BRCA (TCGA) Jun 12012
-Cervix  Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma  CESC (TCGA) Jun 2013
-Colon Colon Adenocarcinoma  COAD (TCGA) Jun 2013
-Head and Neck Head and Neck Squamous Cell Carcinoma HNSC (TCGA) Jun 2013
-Kidney-Chromophobe  Kidney Chromophobe  KICH (TCGA) Jun 2013
-Kidney-Clear-Cell Kidney Renal Clear Cell Carcinoma KIRC (TCGA) Jun 2013
-Kidney-Papillary-Cell Kidney Renal Papillary Cell Carcinoma KIRP (TCGA) Jun 2013
-Liver-Nonviral  Hepatocellular Carcinoma (Secondary to Alcohol and Adiposity) HCCA (ICGC) Mar 2013
-Liver-Viral Hepatocellular Carcinoma (Viral)  HCCV (ICGC) Mar 2013
-Lung-Adenocarcinoma Lung Adenocarcinoma LUAD (TCGA) Jun 2013
-Lung-Squamous Cell  Lung Squamous Cell Carcinoma  LUSC (TCGA) Jun 2013
-Melanoma  Melanoma  ML (Yardena Samuels lab)  Dec 2011
-Other  General purpose  OV (TCGA) Jun 2013
-Ovary Ovarian Serous Cystadenocarcinoma OV (TCGA) Jun 2013
-Pancreas  Pancreatic Cancer PNCC (ICGC))  Mar 2013
-Prostate-Adenocarcinoma Prostate Adenocarcinoma PRAD (TCGA) Jun 2013
-Rectum  Rectum Adenocarcinoma READ (TCGA) Jun 2013
-Skin  Skin Cutaneous Melanoma SKCM (TCGA) Jun 2013
-Stomach Stomach Adenocarcinoma  STAD (TCGA) Jun 2013
-Thyroid Thyroid Carcinoma THCA (TCGA) Jun 2013
-Uterus  Uterine Corpus Endometriod Carcinoma  UCEC (TCGA) Jun 2013
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tool-data/ctat_cravat_tissues.loc.sample	Tue Nov 06 11:53:38 2018 -0500
@@ -0,0 +1,27 @@
+Bladder 	Bladder Urothelial Carcinoma 	BLCA (TCGA) 	Jun 2013
+Blood-Lymphocyte 	Chronic Lymphocytic Leukemia 	CLL (ICGC) 	Mar 2013
+Blood-Myeloid 	Acute Myeloid Leukemia 	LAML (TCGA) 	Jun 2013
+Brain-Cerebellum 	Medulloblastoma 	MB (mixed source) 	Dec 2010
+Brain-Glioblastoma-Multiforme 	Glioblastoma Multiforme 	GBM (TCGA) 	Jun 2013
+Brain-Lower-Grade-Glioma	Brain Lower Grade Glioma	LGG (TCGA)	Jun 2013
+Breast	Breast Invasive Carcinoma	BRCA (TCGA)	Jun 12012
+Cervix	Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma	CESC (TCGA)	Jun 2013
+Colon	Colon Adenocarcinoma	COAD (TCGA)	Jun 2013
+Head and Neck	Head and Neck Squamous Cell Carcinoma	HNSC (TCGA)	Jun 2013
+Kidney-Chromophobe	Kidney Chromophobe	KICH (TCGA)	Jun 2013
+Kidney-Clear-Cell	Kidney Renal Clear Cell Carcinoma	KIRC (TCGA)	Jun 2013
+Kidney-Papillary-Cell	Kidney Renal Papillary Cell Carcinoma	KIRP (TCGA)	Jun 2013
+Liver-Nonviral	Hepatocellular Carcinoma (Secondary to Alcohol and Adiposity)	HCCA (ICGC)	Mar 2013
+Liver-Viral	Hepatocellular Carcinoma (Viral)	HCCV (ICGC)	Mar 2013
+Lung-Adenocarcinoma	Lung Adenocarcinoma	LUAD (TCGA)	Jun 2013
+Lung-Squamous Cell	Lung Squamous Cell Carcinoma	LUSC (TCGA)	Jun 2013
+Melanoma	Melanoma	ML (Yardena Samuels lab)	Dec 2011
+Other	 General purpose	OV (TCGA)	Jun 2013
+Ovary	Ovarian Serous Cystadenocarcinoma	OV (TCGA)	Jun 2013
+Pancreas	Pancreatic Cancer	PNCC (ICGC))	Mar 2013
+Prostate-Adenocarcinoma	Prostate Adenocarcinoma	PRAD (TCGA)	Jun 2013
+Rectum	Rectum Adenocarcinoma	READ (TCGA)	Jun 2013
+Skin	Skin Cutaneous Melanoma	SKCM (TCGA)	Jun 2013
+Stomach	Stomach Adenocarcinoma	STAD (TCGA)	Jun 2013
+Thyroid	Thyroid Carcinoma	THCA (TCGA)	Jun 2013
+Uterus	Uterine Corpus Endometriod Carcinoma	UCEC (TCGA)	Jun 2013
--- a/tool_data_table_conf.xml.sample	Fri Nov 02 14:17:08 2018 -0400
+++ b/tool_data_table_conf.xml.sample	Tue Nov 06 11:53:38 2018 -0500
@@ -3,4 +3,8 @@
         <columns>value, name, path</columns>
         <file path="tool-data/ctat_genome_resource_libs.loc" />
     </table>
+    <table name="ctat_cravat_tissues" comment_char="#" allow_duplicate_entries="False">
+        <columns>value, name, code, date</columns>
+        <file path="tool-data/ctat_cravat_tissues.loc" />
+    </table>
 </tables>