Ondex workflow manual

For Ondex version 0.1.3

Download and installation

1. Download ondex-mini
2. Unzip it to a directory of your choice.

Installing a plugin

1. Choose and download your plugins from the Ondex nexus
2. Copy them into the plugins/ directory inside your ondex-mini installation.
3. Follow any possible additional instructions you may find in the plugins' documentations.

Running a workflow

1. Make sure that all the plugins that are referred to in the workflow have been installed correctly.
2. Execute on the command line:

   ./runme.sh <workflow-xml-file>
   

   where <workflow-xml-file> is the name of the xml file containing your workflow definition.

Composing a workflow file

An Ondex workflow file is an xml file that follows the ondex workflow XSD (XML Schema Definition).

A typical workflow file looks like this:

<?xml version="1.0"?>
<Ondex xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
	xsi:noNamespaceSchemaLocation="ONDEXParameters.xsd">
	<DefaultGraph name="biogrid" type="memory">
		<Parameter name="ReplaceExisting">true</Parameter>
	</DefaultGraph>
	<Parser name="biogrid">
		<Parameter name="ImportFile">data/biogrid/BIOGRID.tab.txt</Parameter>
		<Parameter name="TranslationFile">data/biogrid/translation.tab</Parameter>
	</Parser>
	<Export name="oxl">
		<Parameter name="ExportFile">sce_biogrid_v4.xml.gz</Parameter>
		<Parameter name="GZip">true</Parameter>
	</Export>
</Ondex>

Initialization

You can see that it is a sequence of workflow component invocations, always starting with a graph initialization:

<DefaultGraph name="biogrid" type="memory">
	<Parameter name="ReplaceExisting">true</Parameter>
</DefaultGraph>

Here you have several options:

• You can choose a name for your graph by specifying the name attribute.
• You can determine which graph implementation Ondex will use for this workflow by setting the type attribute accordingly. Currently you can choose from the following implementations:
  • memory: An implementation that will keep the complete graph stored in your RAM.
  • berkeley: An implementation that uses the Berkeley database system to store the graph persistently on your harddrive.

Workflow components

You can invoke any component from the plugins you downloaded by using the tag that corresponds to its type in conjunction with the name attribute that pointing to your processor's id. For example:

<Transformer name="foo"/>

would invoke a transformer called 'foo'.

Here is a list of all component type tag names:

• <Parser/>
• <Mapping/>
• <Filter/>
• <Transformer/>
• <Statistics/>

Arguments and parameters

Each workflow component has a unique set of requirements that is described in its manual.

Parsers and Statistics can require a special attribute specifying an input file or directory, as can be seen in the above example. The attribute names are datafile and datadirectory, respectively.

All other parameters can be passed to the component using Parameter tags, e.g:

<Parser name="biogrid" datafile="data/biogrid/BIOGRID.tab.txt">
	<Parameter name="TranslationFile">data/biogrid/translation.tab</Parameter>
</Parser>

The name attribute corresponds to the parameter's identifier, e.g. 'TranslationFile'. Any value is specified in the space inside the parameter tag.