BasicUssage

1 Previous steps

1.1 Accessing to the UI (User Interface)

• (eg:)ssh -X user@machine
• user is the username
• machine is the name of the UI

1.2 Openning a Grid proxy

• (eg:)voms-proxy-init --voms biomed --valid 50:0
• biomed is the virtual organization (vo)
• --valid 50:0 is an optional flag that shows the expiration time of the proxy (50 hours in this case)
• (eg:)voms-proxy-destroy (To close the current proxy)
• (eg:)voms-proxy-info --all (To consult diverse information about your opened proxy)

1.3 Uploading your images to the UI (User Interface)

• (eg:) scp -r directory user@machine:directory
• directory is the name of the directory that contains your images
• user is the username
• machine is the name of the UI
• directory(2) is the name of the directory that will be created on the user interface

2 Configuring DIANE for xmipp_MLalign2D and xmipp_MLrefine3D

2.1 Adding the paths of GANGA/DIANE and xmipp_MLalign2D/xmipp_MLrefine3D binary files

• The path of GANGA binaries in villon machine is /opt/ganga/install/slc3_gcc323/4.2.1/bin

• The path of DIANE binaries in villon machine is /opt/diane/specific/slc3_gcc323/HEAD/DIANE/dev/commands

• The path of the MLalign2D and MLrefine3D binary files is /opt/xmipp

• (eg:) export PATH=/opt/ganga/install/slc3_gcc323/4.2.1/bin:$PATH

• (eg:) export PATH=/opt/diane/specific/slc3_gcc323/HEAD/DIANE/dev/commands:$PATH

• (eg:) export PATH=/opt/xmipp:$PATH

• You can add the previous lines to the .bashrc file (is placed in your home directory) and run (eg:) source .bashrc - to fix the changes

2.2 DIANE/xmipp_MLalign2D xmipp_MLrefine3D ".job" configuration files

2.2.1 Coping a ".job" file to our home directory

• (eg:) cp /opt/diane/specific/slc3_gcc323/HEAD/DIANE/dev/workspace/xmipp_test.job $HOME

2.2.2 Description of the ".job" file tags

Directory With this variable we indicate the directory in which we have the ".sel" file and the images that it references

• (eg:) dir = '/home/user/imagesdir/' Application With this variable we indicate that we want to use the xmipp_MLalign2D application

• (eg:) Application = 'xmipp' Data (MLalign2D) Data var contains different tags, the description of each one is:

• Program whe can choose between xmipp_MLalign2D and xmipp_MLrefine3D

• Niter contains the number of iteractions

• imgselfile contains the path to the ".sel" file

• inirefsel contains the path to the reference images

• add_params to add more params to the execution

• MLroot Rootname for all output files

• noise not yet implemented

• offset not yet implemented

• Nsplit split the full set of images in different subsets

• iter Number of iteractions of each subset of images (allways at 1)

• (eg:)

Data = {'Niter' : 2,

'Program' : 'MLalign2D'

'imgselfile' : dir+'selfile.sel',

'inirefsel' : dir+'refselfile.sel', # CREATE THE REFERENCE IMAGES FIRST!

'add_params' : '',

'MLroot' : 'output',

'noise' : 0,

'offset' : 0,

'Nsplit' : 3, # SPLIT ALL IMAGES IN A NUMBER OF GROUPS

'iter' : 1 }

Data (MLrefine3D) Data var contains different tags, the description of each one is:

• Program whe can choose between xmipp_MLalign2D and xmipp_MLrefine3D

• Niter contains the number of iteractions

• imgselfile contains the path to the ".sel" file

• inirefsel contains the path to the reference volumes

• add_params to add more params to the execution

• MLroot Rootname for all output files

• Nsplit split the full set of images in different subsets

• iter Number of iteractions of each subset of images (allways at 1)

• (eg:)

Data = {'Niter' : 2,

'Program' : 'MLrefine2D'

'imgselfile' : dir+'selfile.sel',

'inirefsel' : dir+'refselfile.sel', # CREATE THE REFERENCE VOLUMES FIRST!

'add_params' : '',

'MLroot' : 'output',

'Nsplit' : 3, # SPLIT ALL IMAGES IN A NUMBER OF GROUPS

'iter' : 1 }

JobChain It references the Data var

• (eg:) JobChain = [Data] JobInitData Init data for the execution of DIANE/xmipp_MLalign2D jobs

• (eg:) JobInitData = None

• (eg: MLalign2D case) InputFiles = [Data['imgselfile'],Data['inirefsel'],'/opt/xmipp/xmipp_MLalign2D']

• (eg: MLalign3D case) InputFiles = [Data['imgselfile'],Data['inirefsel'],'/opt/xmipp/xmipp_MLrefine3D'] InputFiles Input files needed for the execution of DIANE/xmipp_MLalign2D jobs

• (eg:) InputFiles = [Data['imgselfile'],Data['inirefsel'],'/opt/xmipp/xmipp_MLrefine3D']

• '/opt/xmipp/ is the path of some xmipp programs in our user interface

2.3 Running DIANE with xmipp_MLalign2D and xmipp_MLrefine3D

• tcsh

• source /opt/diane/specific/slc3_gcc323/HEAD/DIANE/etc/environment.csh

• diane.startjob -j $PWD/xmipp_test.job -w5@LCG --ganga Where xmipp_test.job is the file we have defined before

• -w5@ (eg: 5) is the number of workers (CPU's on the Grid), LCG is the backend

• --ganga is the management and job submission tool

• If all goes well you have to obtain something like that: DIANE: 11:43:40: job output in: /home/gcarrera/diane.workspace/jobs/

• You can see the status of your jobs usign the command (eg:)ganga --gui

• When DIANE execution has finished you can consult diverse information accessing to the MonALISA Repository at http://lcgui003.cern.ch:8080/