Mercurial > repos > matnguyen > pathogist

changeset 0:6a73638d707c draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Initial Upload
author matnguyen
date Fri, 16 Nov 2018 15:28:27 -0500
parents
children c1b6f5fbbcad
files tools/concatenator/concatenator tools/concatenator/concatenator.xml tools/consensus_clustering/consensus_clustering.xml tools/correlation_clustering/correlation_clustering.xml tools/distance/distance.xml tools/visualization/visualization.xml
diffstat 6 files changed, 176 insertions(+), 0 deletions(-) [+]
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tools/concatenator/concatenator	Fri Nov 16 15:28:27 2018 -0500
@@ -0,0 +1,21 @@
+#!/usr/bin/env python
+
+import sys
+import argparse as ap
+
+parser = ap.ArgumentParser(prog='concatenater', conflict_handler='resolve',
+                           description="Concatenates all accessions and their associated paths")
+
+input = parser.add_argument_group('Input', '')
+input.add_argument('-n', '--name', nargs='+', required=True, help="Sample name (accessions)")
+input.add_argument('-i', '--input', nargs='+', required=True, help="Paths to calls")
+
+if len(sys.argv) == 1:
+    parser.print_usage()
+    sys.exit(1)
+
+args = parser.parse_args()
+output = open('paths.txt', 'w')
+
+for index,path in enumerate(args.input):
+    output.write("%s=%s\n" % (args.name[index], path))
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tools/concatenator/concatenator.xml	Fri Nov 16 15:28:27 2018 -0500
@@ -0,0 +1,27 @@
+<tool id="concatenator" name="Concatenator" version="1.0.0">
+    <description>creates a list of accessions and the associated path to their variant calls</description>
+    <requirements>
+        <requirement type="package" version="0.0.1">pathogist</requirement>
+    </requirements>
+
+    <command><![CDATA[
+        $__tool_directory__/concatenator -n #for $path in $paths# ${path.element_identifier} #end for#
+        -i #for $path in $paths# $path #end for#
+    ]]></command>
+
+    <inputs>
+        <param name="paths" type="data" label="Collection of calls" help="" optional="False" multiple="True"/>
+    </inputs>
+
+    <outputs>
+        <data name="output" label="Paths of ${on_string}" format="txt" from_work_dir="paths.txt"/>
+    </outputs>
+
+    <help>
+
+    </help>
+
+    <citations>
+
+    </citations>
+</tool>
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tools/consensus_clustering/consensus_clustering.xml	Fri Nov 16 15:28:27 2018 -0500
@@ -0,0 +1,34 @@
+<tool id="consensus_clustering" name="PathOGiST - Consensus Clustering" version="1.0.0">
+    <description>: Performs consensus clustering on multiple clusterings</description>
+    <requirements>
+        <requirement type="package" version="0.0.1">pathogist</requirement>
+    </requirements>
+
+    <command><![CDATA[
+        PATHOGIST consensus $constraints $distances $clusterings $fine_clusterings cc.tsv
+    ]]></command>
+
+    <inputs>
+        <param name="distances" type="data" format="tabular,txt" label="Distance Matrices"
+               help="File contains paths to distance matrices for different clusterings"/>
+        <param name="clusterings" type="data" format="tabular,txt" label="Clusterings"
+               help="File contains paths to clusterings (matrices or lists of clustering assignments)"/>
+        <param name="fine_clusterings" type="data" format="tabular,txt" label="Fine Clusterings"
+               help="File contains names of the clusterings which are the finest"/>
+        <param name="constraints" type="boolean" checked="false" truevalue="--all_constraints" falsevalue=""
+               label="All constraints"
+               help="Add all constraints to the optimization problem, not just those with mixed signs"/>
+    </inputs>
+
+    <outputs>
+        <data name="clustering" label="Consensus clustering on ${on_string}" format="tabular" from_work_dir="cc.tsv"/>
+    </outputs>
+
+    <help>
+
+    </help>
+
+    <citations>
+
+    </citations>
+</tool>
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tools/correlation_clustering/correlation_clustering.xml	Fri Nov 16 15:28:27 2018 -0500
@@ -0,0 +1,31 @@
+<tool id="correlation_clustering" name="PathOGiST - Correlation Clustering" version="1.0.0">
+    <description>: Runs correlation clustering on a distance matrix file</description>
+    <requirements>
+        <requirement type="package" version="0.0.1">pathogist</requirement>
+    </requirements>
+
+    <command><![CDATA[
+        PATHOGIST correlation $constraints $distance $threshold cc.tsv
+    ]]></command>
+
+    <inputs>
+        <param name="distance" type="data" format="tabular,txt" label="Distance Matrix"/>
+        <param name="threshold" type="integer" value="0" label="Threshold Value"
+               help="The threshold value can be determined by visualizing the distance matrix"/>
+        <param name="constraints" type="boolean" checked="false" truevalue="--all_constraints" falsevalue=""
+               label="All Constraints"
+               help="Add all constraints to the optimization problem, not just those with mixed signs"/>
+    </inputs>
+
+    <outputs>
+        <data name="clustering" label="Correlation clustering of ${on_string}" format="tabular" from_work_dir="cc.tsv"/>
+    </outputs>
+
+    <help>
+
+    </help>
+
+    <citations>
+
+    </citations>
+</tool>
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tools/distance/distance.xml	Fri Nov 16 15:28:27 2018 -0500
@@ -0,0 +1,32 @@
+<tool id="distance" name="PathOGiST - Distance" version="1.0.0">
+    <description>: Construct distance matrix from genotyping data</description>
+    <requirements>
+        <requirement type="package" version="0.0.1">pathogist</requirement>
+    </requirements>
+
+    <command><![CDATA[
+        PATHOGIST distance $genotyping $data_type matrix.tsv
+    ]]></command>
+
+    <inputs>
+        <param name="genotyping" type="data" format="tabular,txt" label="Calls"
+               help="File containing paths to signal calls"/>
+        <param name="data_type" type="select" label="Data Type">
+            <option value="MLST" selected="True">MLST</option>
+            <option value="SNP">SNP</option>
+            <option value="CNV">CNV</option>
+        </param>
+    </inputs>
+
+    <outputs>
+        <data name="matrix" label="Distance matrix of ${on_string}" format="tabular" from_work_dir="matrix.tsv"/>
+    </outputs>
+
+    <help>
+
+    </help>
+
+    <citations>
+
+    </citations>
+</tool>
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tools/visualization/visualization.xml	Fri Nov 16 15:28:27 2018 -0500
@@ -0,0 +1,31 @@
+<tool id="visualization" name="PathOGiST - Visualization" version="1.0.0">
+    <description>: Visualize distance matrix</description>
+    <requirements>
+        <requirement type="package" version="0.0.1">pathogist</requirement>
+    </requirements>
+
+    <command><![CDATA[
+        PATHOGIST vis $distance $sample $pdf
+    ]]></command>
+
+    <inputs>
+        <param name="distance" type="data" format="tabular,txt" label="Distance Matrix"/>
+        <param name="sample" type="text" size="30" value="sample" label="Sample Name"/>
+        <param name="pdf" type="boolean" checked="false" truevalue="--save_pdf vis.pdf" falsevalue="" label="Save PDF"/>
+    </inputs>
+
+    <outputs>
+        <data name="diagram" label="Diagram of ${on_string}" format="png" from_work_dir="sample.png"/>
+        <data name="pdf_output" label="PDF of ${on_string}" format="pdf" from_work_dir="vis.pdf">
+            <filter>pdf == True</filter>
+        </data>
+    </outputs>
+
+    <help>
+
+    </help>
+
+    <citations>
+
+    </citations>
+</tool>