Mercurial > repos > galaxyp > custom_pro_db_annotation_data_manager
changeset 9:9fec6c7f46eb draft
planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tools/bumbershoot/custom_pro_db commit 7115cf54f290b51b6a791f9ae288dd907a31fb0a
| author | galaxyp |
|---|---|
| date | Fri, 13 Jan 2017 12:18:39 -0500 |
| parents | f4badf0bf23a |
| children | a0a5aa56d29c |
| files | COPYING data_manager/customProDB_annotation.R data_manager/customProDB_annotation.xml data_manager_conf.xml tool-data/customProDB.loc.sample tool_data_table_conf.xml.sample |
| diffstat | 4 files changed, 192 insertions(+), 187 deletions(-) [+] |
line wrap: on
line diff
--- a/data_manager/customProDB_annotation.R Wed Jun 08 16:00:30 2016 -0400 +++ b/data_manager/customProDB_annotation.R Fri Jan 13 12:18:39 2017 -0500 @@ -1,125 +1,125 @@ -#!/usr/bin/env Rscript - -initial.options <- commandArgs(trailingOnly = FALSE) -script_parent_dir <- dirname(sub("--file=", "", initial.options[grep("--file=", initial.options)])) - -## begin warning handler -withCallingHandlers({ - -library(methods) # Because Rscript does not always do this - -options('useFancyQuotes' = FALSE) - -suppressPackageStartupMessages(library("optparse")) -suppressPackageStartupMessages(library("RGalaxy")) - - -option_list <- list() -option_list$dbkey <- make_option('--dbkey', type='character') -option_list$dbsnp <- make_option('--dbsnp', type='character') -option_list$cosmic <- make_option('--cosmic', type='logical') -option_list$outputFile <- make_option('--outputFile', type='character') -option_list$dbkey_description <- make_option('--dbkey_description', type='character') - -opt <- parse_args(OptionParser(option_list=option_list)) - - -customProDB_annotation <- function( - dbkey = GalaxyCharacterParam(required=TRUE), - dbsnp_str = GalaxyCharacterParam(required=FALSE), - cosmic = GalaxyLogicalParam(required=FALSE), - dbkey_description = GalaxyCharacterParam(required=FALSE), - outputFile = GalaxyOutput("output","json")) -{ - if (!file.exists(outputFile)) - { - gstop("json params file does not exist") - } - - if (length(dbkey_description) < 1) - { - dbkey_description = dbkey - } - - if (length(dbsnp_str) > 0) - { - dbsnp = dbsnp_str - } - else - { - dbsnp = NULL - } - - use_cosmic = NULL - if (length(cosmic) > 0) - { - if (grepl("^hg", dbkey)) - { - use_cosmic = TRUE - } - else - { - gstop("COSMIC annotation requested but dbkey does not indicate a human genome (e.g. hg19)") - } - } - - suppressPackageStartupMessages(library(rjson)) - params = fromJSON(file=outputFile) - target_directory = params$output_data[[1]]$extra_files_path - dir.create(target_directory) - - tryCatch( - { - file.remove(outputFile) - }, error=function(err) - { - gstop("failed to remove json params file after reading") - }) - - ucscTableCodingFastaURL = paste("http://genome.ucsc.edu/cgi-bin/hgTables?db=", dbkey, "&hgSeq.cdsExon=on&hgSeq.granularity=gene&hgSeq.casing=exon&hgSeq.repMasking=lower&hgta_doGenomicDna=get+sequence&hgta_group=genes&hgta_track=refGene&hgta_table=refGene&hgta_regionType=genome", sep="") - ucscTableProteinFastaURL = paste("http://genome.ucsc.edu/cgi-bin/hgTables?db=", dbkey, "&hgta_geneSeqType=protein&hgta_doGenePredSequence=submit&hgta_track=refGene&hgta_table=refGene", sep="") - codingFastaFilepath = paste(target_directory, "/", dbkey, ".cds.fa", sep="") - proteinFastaFilepath = paste(target_directory, "/", dbkey, ".protein.fa", sep="") - - suppressPackageStartupMessages(library(customProDB)) - options(timeout=3600) - - cat(paste("Downloading coding FASTA from:", ucscTableCodingFastaURL, "\n")) - download.file(ucscTableCodingFastaURL, codingFastaFilepath, quiet=T, mode='wb') - - cat(paste("Downloading protein FASTA from:", ucscTableProteinFastaURL, "\n")) - download.file(ucscTableProteinFastaURL, proteinFastaFilepath, quiet=T, mode='wb') - - cat(paste("Preparing Refseq annotation files\n")) - customProDB::PrepareAnnotationRefseq(genome=dbkey, CDSfasta=codingFastaFilepath, pepfasta=proteinFastaFilepath, annotation_path=target_directory, dbsnp=dbsnp, COSMIC=use_cosmic) - - outputPath = paste(dbkey, "/customProDB", sep="") - output = list(data_tables = list()) - output[["data_tables"]][["customProDB"]]=c(path=outputPath, name=dbkey_description, dbkey=dbkey, value=dbkey) - write(toJSON(output), file=outputFile) -} - - -params <- list() -for(param in names(opt)) -{ - if (!param == "help") - params[param] <- opt[param] -} - -setClass("GalaxyRemoteError", contains="character") -wrappedFunction <- function(f) -{ - tryCatch(do.call(f, params), - error=function(e) new("GalaxyRemoteError", conditionMessage(e))) -} - - -suppressPackageStartupMessages(library(RGalaxy)) -do.call(customProDB_annotation, params) - -## end warning handler -}, warning = function(w) { - cat(paste("Warning:", conditionMessage(w), "\n")) - invokeRestart("muffleWarning") -}) +#!/usr/bin/env Rscript + +initial.options <- commandArgs(trailingOnly = FALSE) +script_parent_dir <- dirname(sub("--file=", "", initial.options[grep("--file=", initial.options)])) + +## begin warning handler +withCallingHandlers({ + +library(methods) # Because Rscript does not always do this + +options('useFancyQuotes' = FALSE) + +suppressPackageStartupMessages(library("optparse")) +suppressPackageStartupMessages(library("RGalaxy")) + + +option_list <- list() +option_list$dbkey <- make_option('--dbkey', type='character') +option_list$dbsnp <- make_option('--dbsnp', type='character') +option_list$cosmic <- make_option('--cosmic', type='logical') +option_list$outputFile <- make_option('--outputFile', type='character') +option_list$dbkey_description <- make_option('--dbkey_description', type='character') + +opt <- parse_args(OptionParser(option_list=option_list)) + + +customProDB_annotation <- function( + dbkey = GalaxyCharacterParam(required=TRUE), + dbsnp_str = GalaxyCharacterParam(required=FALSE), + cosmic = GalaxyLogicalParam(required=FALSE), + dbkey_description = GalaxyCharacterParam(required=FALSE), + outputFile = GalaxyOutput("output","json")) +{ + if (!file.exists(outputFile)) + { + gstop("json params file does not exist") + } + + if (length(dbkey_description) < 1) + { + dbkey_description = dbkey + } + + if (length(dbsnp_str) > 0) + { + dbsnp = dbsnp_str + } + else + { + dbsnp = NULL + } + + use_cosmic = FALSE + if (length(cosmic) > 0) + { + if (grepl("^hg", dbkey)) + { + use_cosmic = TRUE + } + else + { + gstop("COSMIC annotation requested but dbkey does not indicate a human genome (e.g. hg19)") + } + } + + suppressPackageStartupMessages(library(rjson)) + params = fromJSON(file=outputFile) + target_directory = params$output_data[[1]]$extra_files_path + dir.create(target_directory) + + tryCatch( + { + file.remove(outputFile) + }, error=function(err) + { + gstop("failed to remove json params file after reading") + }) + + ucscTableCodingFastaURL = paste("http://genome.ucsc.edu/cgi-bin/hgTables?db=", dbkey, "&hgSeq.cdsExon=on&hgSeq.granularity=gene&hgSeq.casing=exon&hgSeq.repMasking=lower&hgta_doGenomicDna=get+sequence&hgta_group=genes&hgta_track=refGene&hgta_table=refGene&hgta_regionType=genome", sep="") + ucscTableProteinFastaURL = paste("http://genome.ucsc.edu/cgi-bin/hgTables?db=", dbkey, "&hgta_geneSeqType=protein&hgta_doGenePredSequence=submit&hgta_track=refGene&hgta_table=refGene", sep="") + codingFastaFilepath = paste(target_directory, "/", dbkey, ".cds.fa", sep="") + proteinFastaFilepath = paste(target_directory, "/", dbkey, ".protein.fa", sep="") + + suppressPackageStartupMessages(library(customProDB)) + options(timeout=3600) + + cat(paste("Downloading coding FASTA from:", ucscTableCodingFastaURL, "\n")) + download.file(ucscTableCodingFastaURL, codingFastaFilepath, quiet=T, mode='wb') + + cat(paste("Downloading protein FASTA from:", ucscTableProteinFastaURL, "\n")) + download.file(ucscTableProteinFastaURL, proteinFastaFilepath, quiet=T, mode='wb') + + cat(paste("Preparing Refseq annotation files\n")) + customProDB::PrepareAnnotationRefseq(genome=dbkey, CDSfasta=codingFastaFilepath, pepfasta=proteinFastaFilepath, annotation_path=target_directory, dbsnp=dbsnp, COSMIC=use_cosmic) + + outputPath = paste(dbkey, "/customProDB", sep="") + output = list(data_tables = list()) + output[["data_tables"]][["customProDB"]]=c(path=outputPath, name=dbkey_description, dbkey=dbkey, value=dbkey) + write(toJSON(output), file=outputFile) +} + + +params <- list() +for(param in names(opt)) +{ + if (!param == "help") + params[param] <- opt[param] +} + +setClass("GalaxyRemoteError", contains="character") +wrappedFunction <- function(f) +{ + tryCatch(do.call(f, params), + error=function(e) new("GalaxyRemoteError", conditionMessage(e))) +} + + +suppressPackageStartupMessages(library(RGalaxy)) +do.call(customProDB_annotation, params) + +## end warning handler +}, warning = function(w) { + cat(paste("Warning:", conditionMessage(w), "\n")) + invokeRestart("muffleWarning") +})
--- a/data_manager/customProDB_annotation.xml Wed Jun 08 16:00:30 2016 -0400 +++ b/data_manager/customProDB_annotation.xml Fri Jan 13 12:18:39 2017 -0500 @@ -1,33 +1,38 @@ -<tool id="custom_pro_db_annotation_data_manager" name="CustomProDB Annotation" tool_type="manage_data" version="0.0.1"> - <description>builder</description> - <command interpreter="Rscript --vanilla">customProDB_annotation.R - --outputFile "${out_file}" - --dbkey "${dbkey}" - --dbsnp "${dbsnp}" - #if $cosmic - --cosmic true - #end if - --dbkey_description ${ dbkey.get_display_text() } - 2>&1 - </command> - <stdio> - <exit_code range=":-1" /> - <exit_code range="1:" /> - </stdio> - <inputs> - <param type="genomebuild" name="dbkey" value="" label="UCSC dbKey for reference genome" /> - <param type="text" name="dbsnp" value="" label="dbSNP identifier currently available from UCSC" help="e.g. 'snp142'" /> - <param type="boolean" name="cosmic" value="" label="Annotate somatic SNPs from COSMIC (human only)" /> - </inputs> - <outputs> - <data name="out_file" format="data_manager_json"/> - </outputs> - - <help> - -.. class:: infomark - -**Notice:** If you leave name, description, or id blank, it will be generated automatically. - - </help> -</tool> +<tool id="custom_pro_db_annotation_data_manager" name="CustomProDB Annotation" tool_type="manage_data" version="0.0.1"> + <description>builder</description> + <command interpreter="Rscript --vanilla">customProDB_annotation.R + --outputFile "${out_file}" + --dbkey "${dbkey}" + --dbsnp "${dbsnp}" + #if $cosmic + --cosmic true + #end if + --dbkey_description ${ dbkey.get_display_text() } + 2>&1 + </command> + <stdio> + <exit_code range=":-1" /> + <exit_code range="1:" /> + </stdio> + <inputs> + <param type="genomebuild" name="dbkey" value="" label="UCSC dbKey for reference genome" /> + <param type="text" name="dbsnp" value="" label="dbSNP identifier currently available from UCSC" help="e.g. 'snp142'" /> + <param type="boolean" name="cosmic" value="" label="Annotate somatic SNPs from COSMIC (human only)" /> + </inputs> + <outputs> + <data name="out_file" format="data_manager_json"/> + </outputs> + + <help> + +.. class:: infomark + +**Notice:** If you leave name, description, or id blank, it will be generated automatically. + + </help> + <citations> + <citation type="doi">10.1093/bioinformatics/btt543</citation> + <citation type="bibtex">@misc{toolsGalaxyP, author = {Chambers MC, et al.}, title = {Galaxy Proteomics Tools}, publisher = {GitHub}, journal = {GitHub repository}, + year = {2017}, url = {https://github.com/galaxyproteomics/tools-galaxyp}}</citation> <!-- TODO: fix substitution of commit ", commit = {$sha1$}" --> + </citations> +</tool>
--- a/data_manager_conf.xml Wed Jun 08 16:00:30 2016 -0400 +++ b/data_manager_conf.xml Fri Jan 13 12:18:39 2017 -0500 @@ -1,22 +1,22 @@ -<?xml version="1.0"?> -<data_managers> - - <data_manager tool_file="data_manager/customProDB_annotation.xml" id="custom_pro_db_annotation_builder" version="0.0.1"> - <data_table name="customProDB"> - <output> - <column name="value" /> - <column name="dbkey" /> - <column name="name" /> - <column name="path" output_ref="out_file" > - <move type="directory" relativize_symlinks="True"> - <!-- <source>${path}</source>--> <!-- out_file.extra_files_path is used as base by default --> <!-- if no source, eg for type=directory, then refers to base --> - <target base="${GALAXY_DATA_MANAGER_DATA_PATH}">${dbkey}/customProDB</target> - </move> - <value_translation>${GALAXY_DATA_MANAGER_DATA_PATH}/${dbkey}/customProDB</value_translation> - <value_translation type="function">abspath</value_translation> - </column> - </output> - </data_table> - </data_manager> - -</data_managers> +<?xml version="1.0"?> +<data_managers> + + <data_manager tool_file="data_manager/customProDB_annotation.xml" id="custom_pro_db_annotation_builder" version="0.0.1"> + <data_table name="customProDB"> + <output> + <column name="value" /> + <column name="dbkey" /> + <column name="name" /> + <column name="path" output_ref="out_file" > + <move type="directory" relativize_symlinks="True"> + <!-- <source>${path}</source>--> <!-- out_file.extra_files_path is used as base by default --> <!-- if no source, eg for type=directory, then refers to base --> + <target base="${GALAXY_DATA_MANAGER_DATA_PATH}">${dbkey}/customProDB</target> + </move> + <value_translation>${GALAXY_DATA_MANAGER_DATA_PATH}/${dbkey}/customProDB</value_translation> + <value_translation type="function">abspath</value_translation> + </column> + </output> + </data_table> + </data_manager> + +</data_managers>
--- a/tool_data_table_conf.xml.sample Wed Jun 08 16:00:30 2016 -0400 +++ b/tool_data_table_conf.xml.sample Fri Jan 13 12:18:39 2017 -0500 @@ -1,7 +1,7 @@ -<tables> - <!-- Locations of all customProDB annotations under genome directory --> - <table name="customProDB" comment_char="#"> - <columns>value, dbkey, name, path</columns> - <file path="tool-data/customProDB.loc" /> - </table> -</tables> +<tables> + <!-- Locations of all customProDB annotations under genome directory --> + <table name="customProDB" comment_char="#"> + <columns>value, dbkey, name, path</columns> + <file path="tool-data/customProDB.loc" /> + </table> +</tables>