Mercurial > repos > bgruening > blockclust
changeset 0:5365d08bead0 draft default tip
planemo upload for repository https://github.com/bgruening/galaxytools/tree/master/tools/rna_tools/blockclust commit 1973f3035c10db80883d80847ea254289f5cce2a-dirty
| author | bgruening |
|---|---|
| date | Thu, 17 Sep 2015 16:49:52 -0400 |
| parents | |
| children | |
| files | blockclust.tar.gz blockclust.xml tool_dependencies.xml |
| diffstat | 3 files changed, 319 insertions(+), 0 deletions(-) [+] |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/blockclust.xml Thu Sep 17 16:49:52 2015 -0400 @@ -0,0 +1,215 @@ +<tool id="blockclust" name="BlockClust" version="1.0.0"> + <description>efficient clustering and classification of non-coding RNAs from short read RNA-seq profiles</description> + <requirements> + <requirement type="package" version="1.0">blockclust</requirement> + <requirement type="package" version="1.1">eden</requirement> + <requirement type="package" version="3.0.3">R</requirement> + <requirement type="package" version="0.1.19">samtools</requirement> + <requirement type="package" version="12.135">mcl</requirement> + <requirement type="package" version="1.0">blockclust_rlibs</requirement> + </requirements> + <version_command>echo '1.0'</version_command> + <command> +<![CDATA[ + #if str($tool_mode.operation) == "pre": + BlockClustPipeLine.pl -m PRE -bam $tool_mode.reads_bam -tbed $tags_bed + #elif str($tool_mode.operation) == "clust": + #set $outputdir = $clusters.files_path + #set $accept_bed=list() + #set $reject_bed=list() + ## prepare annotations + #if str($tool_mode.reference) == "hg19": + $accept_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/hg19/hg19.accept.bed") + $reject_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/hg19/hg19.reject.bed") + #elif str($tool_mode.reference) == "mm10": + $accept_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/mm10/mm10.accept.bed") + $reject_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/mm10/mm10.reject.bed") + #elif str($tool_mode.reference) == "dm3": + $accept_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/dm3/dm3.accept.bed") + $reject_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/dm3/dm3.reject.bed") + #elif str($tool_mode.reference) == "rheMac3": + $accept_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/rheMac3/rheMac3.accept.bed") + $reject_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/rheMac3/rheMac3.reject.bed") + #elif str($tool_mode.reference) == "panTro4": + $accept_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/panTro3/panTro4.accept.bed") + $reject_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/panTro3/panTro4.reject.bed") + #elif str($tool_mode.reference) == "xenTro3": + $accept_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/xenTro3/xenTro3.accept.bed") + $reject_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/xenTro3/xenTro3.reject.bed") + #elif str($tool_mode.reference) == "celWS235": + $accept_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/celWS235/celWS235.accept.bed") + $reject_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/celWS235/celWS235.reject.bed") + #elif str($tool_mode.reference) == "tair10": + $accept_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/tair10/tair10.accept.bed") + $reject_bed.append("\$BLOCKCLUST_DATA_PATH/annotations/tair10/tair10.reject.bed") + #end if + BlockClustPipeLine.pl -m TEST -c \$BLOCKCLUST_DATA_PATH/blockclust.config + -t $tool_mode.input_bbo + -a #echo ''.join( $accept_bed ) + -r #echo ''.join( $reject_bed ) + -o $outputdir + #if $tool_mode.nochr: + -nochr + #end if + #if str($tool_mode.pred.enable_pred) == "yes": + -p + -pm $tool_mode.pred.pred_mode + -md \$BLOCKCLUST_DATA_PATH/models; + #if str($tool_mode.pred.pred_mode) == "nearest_neighbour": + cp #echo os.path.join($outputdir,'nearest_neighbour_predictions.txt')# $nearest_neighbour_pred_bed; + #elif str($tool_mode.pred.pred_mode) == "model_based": + cp #echo os.path.join($outputdir,'model_based_predictions.txt')# $model_based_pred_bed; + #end if + #else: + ; + #end if + + cp #echo os.path.join($outputdir, 'mcl_clusters','all_clusters.bed')# $clusters; + cp #echo os.path.join($outputdir, 'hclust_tree.pdf')# $hclust_plot; + cp #echo os.path.join($outputdir, 'discretized.gspan.tab')# $sim_tab_out + #elif str($tool_mode.operation) == "post": + BlockClustPipeLine.pl -m POST -cbed $tool_mode.clusters_bed -cm $tool_mode.cmsearch_out -tab $tool_mode.sim_tab_in -rfam \$BLOCKCLUST_DATA_PATH/rfam_map.txt -o ./; + #end if +]]> + </command> + <inputs> + <conditional name="tool_mode"> + <param name="operation" type="select" label="Select mode of operation"> + <option value="pre">Pre-processing </option> + <option value="clust">Clustering and classification</option> + <option value="post">Post-processing</option> + </param> + <when value="pre"> + <param name="reads_bam" type="data" format="bam" label="BAM file containing alignments" /> + </when> + <when value="clust"> + <param name="input_bbo" type="data" format="tabular" label="Input blockgroups file" /> + <param name="reference" type="select" label="Select reference genome"> + <option value="hg19">Human (hg19)</option> + <option value="mm10">Mouse (mm10)</option> + <option value="dm3">Fly (dm3)</option> + <option value="rheMac3">Monkey (rheMac3)</option> + <option value="panTro4">Chimp (panTro4)</option> + <option value="xenTro3">Frog (xenTro3)</option> + <option value="celWS235">C. elegans (celWS235)</option> + <option value="tair10">Arabidopsis thaliana (tair10)</option> + </param> + <param name="nochr" type="boolean" label="My input files have no 'chr' for chromosome names" checked="False"/> + <conditional name="pred"> + <param name="enable_pred" type="select" label="Would you like to perform classification?"> + <option value="no">No</option> + <option value="yes">Yes</option> + </param> + <when value="yes"> + <param name="pred_mode" type="select" label="Mode of classification"> + <option value="model_based">Model based</option> + <option value="nearest_neighbour">Nearest neighbour</option> + </param> + </when> + </conditional> + </when> + <when value="post"> + <param name="cmsearch_out" type="data" format="tabular" label="Output of cmsearch tool" /> + <param name="clusters_bed" type="data" format="bed" label="BED file containing clusters (output of BlockClust)" /> + <param name="sim_tab_in" type="data" format="tabular" label="Pairwise similarities file" /> + </when> + </conditional> + </inputs> + + <outputs> + <data format="bed" name="tags_bed" label="BlockClust: BAM to BED on ${on_string}"> + <filter> tool_mode["operation"]=="pre"</filter> + </data> + <data format="pdf" name="hclust_plot" label="BlockClust: Hierarchical clustering plot on ${on_string}" > + <filter> tool_mode["operation"]=="clust"</filter> + </data> + <data format="bed" name="clusters" label="BlockClust: BED of predicted clusters on ${on_string}"> + <filter> tool_mode["operation"]=="clust"</filter> + </data> + <data format="bed" name="model_based_pred_bed" label="BlockClust: Model based predictions BED on ${on_string}"> + <filter> + (( + tool_mode["operation"] == 'clust' and + tool_mode["pred"]["enable_pred"] == "yes" and + tool_mode["pred"]["pred_mode"] == "model_based" + )) + </filter> + </data> + <data format="bed" name="nearest_neighbour_pred_bed" label="BlockClust: Nearest neighbor predictions BED on ${on_string}"> + <filter> + (( + tool_mode["operation"] == 'clust' and + tool_mode["pred"]["enable_pred"] == "yes" and + tool_mode["pred"]["pred_mode"] == "nearest_neighbour" + )) + </filter> + </data> + <data format="tabular" name="sim_tab_out" label="BlockClust: Pairwise similarities on ${on_string}"> + <filter> tool_mode["operation"]=="clust"</filter> + </data> + <data format="pdf" name="cluster_dist" from_work_dir="cluster_distribution.pdf" label="BlockClust: Cluster distribution on ${on_string}" > + <filter> tool_mode["operation"]=="post"</filter> + </data> + <data format="pdf" name="cluster_hclust" from_work_dir="hclust_tree_clusters.pdf" label="BlockClust: Hierarchical clustering plot of cluster centroids on ${on_string}" > + <filter> tool_mode["operation"]=="post"</filter> + </data> + </outputs> + <help> +<![CDATA[ + +**What it does** + +BlockClust is an efficient approach to detect transcripts with similar +processing patterns. BlockClust proposes a novel way to encode expression profiles +in compact discrete structures which can then be processed using +fast graph-kernel techniques. BlockClust allows both clustering and +classification of small non-coding RNAs. + +BlockClust runs in three operating modes: + +1) Pre-processing - converts given mapped reads (BAM) into BED file of tags + +2) Clustering and classification - of given input blockgroups (output of blockbuster tool) as explained in the original paper. + +3) Post-processing - plots for overview of predicted clusters. + +For a thorough analysis of your data, we suggest you to use complete blockclust workflow, which contains all three modes of operation. + +**Input** + +BlockClust input files are dependent on the mode of operation: + +1. Pre-processing mode: + * Binary Sequence Alignment Map (BAM) file + +2. Clustering and classification: + * A blockgroups file generated by blockbuster tool + * Select reference genome + +3. Post-processing: + * Output of cmsearch, searched clusters generated by BlockClust against Rfam + * BED file containing clusters generated by BlockClust + * Pairwise similarities of blockgroups generated by BlockClust + +**Output** + +1. Pre-processing mode: + * BED file of tags with expressions + +2. Clustering and classification: + * Hierarchical clustering plot of all input blockgroups by their similarity + * Pairwise similarities of all input blockgroups + * BED file containing predicted clusters + * BED file containing prediction of blockgroups by pre-compiled SVM binary classification model. + +3. Post-processing: + * Plot of distribution of ncRNA families per predicted cluster (overview of cluster precissions). The annotation of ncRNA families are retrieved by searching cluster instances against Rfam database. + * Hierarchical clustering made out of centroids of each BlockClust predicted cluster + + +]]> + </help> + <citations> + <citation type="doi">10.1093/bioinformatics/btu270</citation> + </citations> +</tool>
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/tool_dependencies.xml Thu Sep 17 16:49:52 2015 -0400 @@ -0,0 +1,104 @@ +<?xml version="1.0"?> +<tool_dependency> + <package name="samtools" version="0.1.19"> + <repository changeset_revision="0e56e4dac6e7" name="package_samtools_0_1_19" owner="iuc" toolshed="https://toolshed.g2.bx.psu.edu" /> + </package> + <package name="R" version="3.0.3"> + <repository changeset_revision="6ce19fda3a9a" name="package_r_3_0_3" owner="iuc" toolshed="https://toolshed.g2.bx.psu.edu" /> + </package> + <package name="eden" version="1.1"> + <repository changeset_revision="767f6e9800dd" name="package_eden_1_1" owner="rnateam" toolshed="https://toolshed.g2.bx.psu.edu" /> + </package> + <package name="mcl" version="12.135"> + <repository changeset_revision="7003196be1b1" name="package_mcl_12_135" owner="iuc" toolshed="https://toolshed.g2.bx.psu.edu" /> + </package> + <package name="blockclust_rlibs" version="1.0"> + <install version="1.0"> + <actions> + <action type="setup_r_environment"> + <repository changeset_revision="6ce19fda3a9a" name="package_r_3_0_3" owner="iuc" toolshed="https://toolshed.g2.bx.psu.edu"> + <package name="R" version="3.0.3" /> + </repository> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/ape_3.0-11.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/plyr_1.8.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/digest_0.6.4.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/gtable_0.1.2.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/dichromat_2.0-0.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/colorspace_1.2-4.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/munsell_0.4.2.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/labeling_0.2.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/RColorBrewer_1.0-5.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/stringr_0.6.2.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/scales_0.2.3.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/proto_0.3-10.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/reshape2_1.2.2.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/ggplot2_0.9.3.1.tar.gz</package> + <package>https://github.com/bgruening/download_store/raw/master/blockclust/r-packages/squash_1.0.1.tar.gz</package> + </action> + </actions> + </install> + <readme> + + </readme> + </package> + <package name="blockclust" version="1.0"> + <install version="1.0"> + <actions> + <action type="download_by_url">https://github.com/bgruening/download_store/raw/master/blockclust/blockclust-1.0.tar.gz</action> + <action type="shell_command">make</action> + <action type="move_file"> + <source>BlockClust</source> + <destination>$INSTALL_DIR/bin</destination> + </action> + <action type="move_file"> + <source>BlockClustPipeLine.pl</source> + <destination>$INSTALL_DIR/bin</destination> + </action> + <action type="move_file"> + <source>filterKnownBlockGroups.pl</source> + <destination>$INSTALL_DIR/bin</destination> + </action> + <action type="move_file"> + <source>perf</source> + <destination>$INSTALL_DIR/bin</destination> + </action> + <action type="move_file"> + <source>plotClusters.R</source> + <destination>$INSTALL_DIR/bin</destination> + </action> + <action type="move_file"> + <source>blockclust.config</source> + <destination>$INSTALL_DIR/data/</destination> + </action> + <action type="move_file"> + <source>rfam_map.txt</source> + <destination>$INSTALL_DIR/data/</destination> + </action> + <action type="make_directory">$INSTALL_DIR/data/annotations</action> + <action type="download_file">https://github.com/bgruening/download_store/raw/master/blockclust/blockclust-data-1.0/annotations/dm3.tar.gz</action> + <action type="shell_command">tar xfvz dm3.tar.gz -C $INSTALL_DIR/data/annotations/</action> + <action type="download_file">https://github.com/bgruening/download_store/raw/master/blockclust/blockclust-data-1.0/annotations/hg19.tar.gz</action> + <action type="shell_command">tar xfvz hg19.tar.gz -C $INSTALL_DIR/data/annotations/</action> + <action type="download_file">https://github.com/bgruening/download_store/raw/master/blockclust/blockclust-data-1.0/annotations/mm10.tar.gz</action> + <action type="shell_command">tar xfvz mm10.tar.gz -C $INSTALL_DIR/data/annotations/</action> + <action type="download_file">https://github.com/bgruening/download_store/raw/master/blockclust/blockclust-data-1.0/annotations/panTro4.tar.gz</action> + <action type="shell_command">tar xfvz panTro4.tar.gz -C $INSTALL_DIR/data/annotations/</action> + <action type="download_file">https://github.com/bgruening/download_store/raw/master/blockclust/blockclust-data-1.0/annotations/rheMac3.tar.gz</action> + <action type="shell_command">tar xfvz rheMac3.tar.gz -C $INSTALL_DIR/data/annotations/</action> + <action type="download_file">https://github.com/bgruening/download_store/raw/master/blockclust/blockclust-data-1.0/annotations/celWS235.tar.gz</action> + <action type="shell_command">tar xfvz celWS235.tar.gz -C $INSTALL_DIR/data/annotations/</action> + <action type="download_file">https://github.com/bgruening/download_store/raw/master/blockclust/blockclust-data-1.0/annotations/tair10.tar.gz</action> + <action type="shell_command">tar xfvz tair10.tar.gz -C $INSTALL_DIR/data/annotations/</action> + <action type="download_file">https://github.com/bgruening/download_store/raw/master/blockclust/blockclust-data-1.0/models.tar.gz</action> + <action type="shell_command">tar xfvz models.tar.gz -C $INSTALL_DIR/data/</action> + <action type="set_environment"> + <environment_variable action="prepend_to" name="PATH">$INSTALL_DIR/bin</environment_variable> + <environment_variable action="set_to" name="BLOCKCLUST_DATA_PATH">$INSTALL_DIR/data</environment_variable> + </action> + </actions> + </install> + <readme> + + </readme> + </package> +</tool_dependency>