Taverna Workflows

1
Taverna Workflows A beginners guide
Mark Wilkinson Edward Kawas (Modified for
Taverna 2.1 from earlier version by Paul Fisher)

2
Taverna
This tutorial is designed to introduce you to the
Taverna 2.1 workflow workbench
3
1. Installing the Workbench
4
Prerequisites - 1

• Java
• In order to run Taverna 2.1 on your computer you
  will need to have the latest Java installed. If
  you do not have Java already installed, you can
  download it from this URL
• http//java.sun.com/javase/downloads/index.jsp
• You will have a choice of the download you would
  like. Download the JDK with Java EE packaged up
  too. This will give you the opportunity to
  develop web services and use the ones deployed by
  Java developers at a later date. The Java Runtime
  Environment (JRE) being downloaded should be 1.6
  or later for Taverna to work.
• If you have Java installed, but it is an earlier
  version, you will need to update it to 1.6 or
  later otherwise Taverna will NOT work.
• The minimal installation you will need is the
  standard JDK package.
• Download the desired JDK by following the link on
  the website and choose a location on your
  computer to save it to.
• Open the saved file and follow the installation
  instructions to install Java on your computer
• Restart your computer to complete the
  installation.

5
Prerequisites - 2

• A zip package (not needed for MS Windows install)
• You will also need a tool to unzip the downloaded
  workbench. There are various tools available on
  the internet, including WinZip, 7-Zip, and a few
  others. Personally I prefer 7-Zip, which is free
  to and easy use, available at the following URL
• http//www.7-zip.org/download.html
• You will need to choose the appropriate file to
  download for your operating system, i.e. Windows,
  Linux, Apple MAC.
• Choose a location to save the file in and save
  it.
• Locate your saved file and follow the
  installation instructions to install it on your
  computer.
• Restart your computer to complete the
  installation.

6
Prerequisites - 3

• Linux users - Graphviz
• Those who are installing Taverna on Linux will
  also have to install Graphviz onto the system.
  This is available at the following URL
• http//www.graphviz.org/
• At the time of writing I have no installation
  instructions for this package, so please refer to
  the user documentation provided on the web site

7
Downloading Taverna

• Updates and news about Taverna can be found on
  the myGrid homepage at
• http//www.mygrid.org.uk
• The latest versions of Taverna can be downloaded
  from Sourceforge
• http//taverna.sourceforge.net
• Once on the Download page, identify the
  relevant Taverna distribution you need. (we are
  using 2.1)
• Follow the link to download the workbench. The
  web page should re-direct you to the source forge
  page.
• Choose a location to save the file and click OK.

8
Unzipping the workbench(not needed for a Windows
installation)

• Choose to Unzip/Extract the files, but not into
  the current directory.
• You will need to choose a directory in which to
  unzip the files. I recommend somewhere in the
  root drive of your computer so you can easily
  access it, e.g. C\myGrid\ .
• You can change the name of the folder at this
  stage, e.g. to Taverna.
• If you are using Taverna on Linux, please be sure
  that you have the relevant access permissions to
  install and run Taverna in the desired directory.

If you need a Zip package download and install
7-ZIP (find it using Google)
9
Opening Taverna

• Locate you Taverna installation in the windows
  Start menu, or open the Taverna folder.
• Start Taverna by double clicking on the Taverna
  icon in the start menu, or runme.sh (Linux and
  Mac users).
• If you have successfully installed Java, you
  should see a dialog box or command window open,
  shortly followed by the Taverna application.
• Once you have installed Taverna for the first
  time it will need to update all of its
  components. You do not need to do anything for
  this, as this happens as the workbench is
  opening. You should see a graphic in the centre
  of your screen, with a download progress. Each
  component will be shown loading in this progress
  bar in turn. Once this has completed (depending
  on connection speed about 5 minutes), the
  Taverna workbench will open.
• The Taverna workbench consists of 3 main panels
  for constructing workflows
• The Available services pane (Top Left side)
• The Advanced Model Explorer pane (Bottom Left
  side)
• The Diagram pane (Right side)

10
(No Transcript)
11
The 3 Panes of Taverna

• The Available services pane is used to display
  the web services to the user. This list contains
  default services from when the workbench starts.
  Once you become more experienced with the
  workbench, you will be able to add you own
  services, including adding default services so
  they load automatically when Taverna opens. This
  list contains WSDL web services, local BioJava
  widgets, Soaplab services, and BioMoby objects.
  Each of these can be added to the workflow model
  (workflow being constructed) so that a task can
  be achieved.
• The Advanced Model Explorer (AME) pane contains
  the services used in the current workflow,
  including the inputs, outputs, and data links
  between each service. Once populated with
  services, each service can be expanded using the
  button. This provides a list of the inputs
  and outputs that the service takes in and expels.
  It is these inputs and outputs that allow you to
  connect services together.
• The Diagram pane shows a graphical representation
  of the workflow being used/constructed. The
  diagram can be adapted to view different aspects
  of the current workflow, to show all the ports
  for all the services, only those ports that have
  been connected or bound, or to change the layout
  of the workflow from portrait to landscape.

12
3 Panes of Taverna
Available services
Available services
Available services
Diagram pane
Diagram pane
Diagram pane
Advanced Model Explorer
Advanced Model Explorer
13
Advanced Model Explorer

• AME (bottom left panel)
• The AME is the primary editing component within
  Taverna. Through it you can load, save, and edit
  any property of a workflow.
• It enables you to
• build a workflow
• add nested workflows
• edit workflows by connecting services
• add metadata to a workflow

14
Diagram Pane

• Shows inputs / outputs, services and control
  flows
• It allows you to change the view of a workflow,
  save the visual representation, and explode or
  implode nested workflows

15
Available services

• Lists services available by default in Taverna
  top left
• 3500 services
• Local java services
• Simple web services
• Soaplab services legacy command-line
  application
• R Processor
• BioMart database services
• BioMoby services
• Beanshell processor
• Allows the user to add new services or workflows
  from the web or from file systems

16
2. Adding new services
17
Exercise 2 Adding New Services

• New services can be gathered from anywhere on the
  web the default list are just a few we already
  know about importing others is very
  straightforward
• Go to the DDBJ list of available web services at
  http//xml.nig.ac.jp/index.html
• These services were not designed for use in
  Taverna, but Taverna can use them if you supply
  the address of the WSDL file
• Click on the DDBJ blast service
  (http//xml.nig.ac.jp/wsdl/Blast.wsdl) and copy
  the web page address

18
This is the Blast Service from DDBJ
19
This is what Taverna needs
Clicking on the blast service brings you to the
page describing the methods(functions) that
the blast service can execute
20
Exercise 2 Adding New Services

• Go to the Service panel in Taverna, and click on
  Import new services button. You are then asked
  what kind of service you are adding. The DDBJ
  services are WSDL so we will select WSDL
  service from the menu.
• A window will pop-up asking for a web address
• Enter the Blast Web service address you just
  copied
• Scroll down to the bottom of the Services list
  and look at the new DDBJ service that is now
  included, clicking on the icon next to the
  service

21
Exercise 2 Adding a WSDL service
22
Exercise 2 Adding New Services

• Enter the Blast Web service address you just
  copied

23
Exercise 2 Adding New Services

• Scroll down to the bottom of the Services list
  and look at the new DDBJ service that is now
  included, clicking on the icon next to the
  service

There it is!
24
3. Finding services and adding them to workflows
25
Exercise 3 - Finding services and adding them to
workflows

• We want to do a BLAST search using the DDBJ
  service that we just added into Taverna
• First, we need to get a sequence to blast, and
  this sequence should be in FASTA format.
• Go to the Services Panel
• Type fasta into the search box at the top of
  the panel You will see several services, the top
  one sounds promising!

26
Exercise 3 - Finding services and adding them to
workflows
Right-click andadd to workfow
27
Exercise 3 - Finding services and adding them to
workflows
28
Exercise 3 - Finding services and adding them to
workflows
To see the inputs and outputs for a service,
click the little button here
29
Exercise 3 - Finding services and adding them to
workflows

• Now we need to provide it an ID number for the
  sequence we are interested in.
• Right-click on the id box in the service
  diagram and set constant value (e.g. 163483)

30
Exercise 3 - Finding services and adding them to
workflows
31
Exercise 3 - Finding services and adding them to
workflows

• Now we want to add the Blast service from DDBJ
• Search for blast, or simply scroll down to the
  DDBJ service searchSimple execute blast

32
Exercise 3 - Finding services and adding them to
workflows

• Right-click and add it to the workflow

33
Exercise 3 - Finding services and adding them to
workflows

• Now we need to connect the output from the
  sequence retrieval to the input port of the blast
• Right click on outputText connect as input
  to, searchSimple, query
• How do we know the correct input port was called
  query? In this case we can get the
  documentation from the DDBJ website by clicking
  on the searchSimple link
• Unfortunately, not all services are so nicely
  documented!!
• (See the BioMoby project for a solution to this
  problem! http//biomoby.org)

34
Documentation for the DDBJ service
35
Exercise 3 - Finding services and adding them to
workflows
36
Exercise 3 - Finding services and adding them to
workflows
37
Exercise 3 - Finding services and adding them to
workflows

• The other two ports are used to name the database
  you want to blast against, and the type of blast
  you want to do
• Lets chose to execute a blastn against the
  Arabidopsis EST dataset
• Right-click on database and set constant
  value. Make the value est_atha
• Right-click on program, set constant value
  and make the value blastn

38
Exercise 3 - Finding services and adding them to
workflows
39
Exercise 3 - Finding services and adding them to
workflows

• Finally we need to do something with the output.
  Lets add an output bin to store it
• Right click on Result from the blast service,
  and connect as output to new workflow output
  port

40
DONE!!
41
Running your workflow

• In the main menu bar, there is a play button
  to run your workflow, click it!

42
Running your workflow

• Services become grey as they finish executing.
  Once all services are grey, your workflow has
  completed

43
Running your workflow

• To see the results, click on the one of the
  result links on the left. If the workflow
  generated more than one result, there will be
  multiple links in that list.

Click here
Result here!
44
Lets make it more flexible!

• We dont want to have to re-set the constant
  input value every time, so lets make it more
  flexible. Right-click on the id_value and
  delete it.
• Now right-click on the Get_nucleotide_fasta id
  port and connect with output from new workflow
  input port

45
Lets make it more flexible!

• Lets say we had a lot of sequences we wanted to
  blast make your input port into a list!

46
Your new, flexible workflow
47
Running the workflow

• This time, when you press play, you will be
  asked for input

48
Running the workflow

• Click on new value for each of your list of
  genbank gi numbers

Click here to begin
49
Done!
Three blast reports