Upgrading the Proline Suite

From Beta 1 to Beta 2:

To update the differents Proline Databases, you need to get the appropriated version of ProlineAdmin. ProlineAdmin is released as an archive file, extract this archive in your preferred folder to install this application. You can also use the provided graphical tool, ProlineAdmin GUI.

The configuration files didn't change. You can just copy theses files from your previous ProlineAdmin to the new one. Configuration files are locates under config folder. Note: default naming scheme of databases created by Proline can be modified by editing config/application-<dbtype>.conf file. (To see configuration description, see first install section.)

Update datastore by executing the following command:

> run_cmd.bat upgrade_dbs

or, on linux system

> run_cmd.sh upgrade_dbs

Open the graphical interface by running start.bat (or start.sh on linux system). If you used the Windows installer, a shortcut “Proline Admin” is available in the Windows Start Menu, under the Proline folder.

If you don't have a valid configuration file, you can change it or edit it (see how to).

Click the button named Upgrade Proline databases.

Note: This functionnality will be disable if the configuration is invalid or if Proline is not set up (see how to configure and set up Proline, or using ProlineAdmin GUI)

Wait until this operation is finished, it may be relatively long if lots of result files with spectral matches have been imported !

Download apropriated Proline WebCore and extract the resulting zip file to the location of your choice, use a different location than previous installation. Before running Jetty, you first need to modify ProlineWeb-Core configuration.

The configuration files, located at <your_folder>\ProlineWeb-Core-x.y.z\Proline\WEB-INF\classes, are different from previous version.

You can copy the datastore description from previous release back to the new application.conf file (see Setting up the Datastore).

You also need to configure the mount_points entry as explain here

Result identification files (Mascot dat, OMSSA or X!Tandem) as well as mzDB files (for the XIC Quantitation process) are only browsed from Proline Server side.

Administrator must configure the target directory(ies) in the entry mount_points in the application.conf file by editing the file manually or using ProlineAdmin GUI.

Mascot dat or OMSSA path should be configured in result_files sub-entry, administrator can add one or more mappings as label = “<absolute/directory/path>”. mzDB files path should be set under mzdb_files sub-entry.

Label can be any valid string chosen by Administrator to help user identify mount_point. If multiple repositories are defined, labels must be different.

Configuration examples :

mount_points {

  result_files {
    mascot_data = "Z:/"
    omssa_data = "/local/omssa/data"
    xtandem_data = "U:/"
    ...
  }
  ...
  
  mzdb_files {
    mzd_path = "/local/data/mzdb"
  }

}

Run jetty-runner.bat (or jetty-runner.sh on linux system) to start the jetty server. You should now be able to access ProlineWeb-Core by typing http://localhost:8080/proline orhttp://<host>:8080/proline in your favorite browser. The following message must appear :

ProlineWeb-Core working !

Number of IVersion services : <X>
fr.proline.core.wsl.Version Module: ProlineWeb-Core Version: <XXX>
fr.proline.module.parser.omssa.Version Module: PM-OmssaParser Version: <YYY>
fr.proline.module.parser.mascot.Version Module: PM-MascotParser Version:<XYZ>
fr.proline.admin.Version Module: Proline-Admin Version: <ZYW>
fr.proline.util.ScalaVersion Module: ProFI-Commons-Scala Version: <YZX>
fr.proline.util.JavaVersion Module: ProFI-Commons-Java Version: <YXZ>
fr.proline.core.service.Version Module: Proline-OMP Version: <WYZ>

In previous version, there were a common logging between Proline Web and Prolibe Server but different passwords. We changed this behaviour, so after you upgrade your proline server, your password will be reset to 'proline'. You can change this from Proline Studio (Connection Managment page) or Proline Web

A new module is available in Proline Beta2. Even if this is a optionnal module it is recommended to install it, mostly if you want to view the proteins sequences in the user interfaces !

This module is distributed as an archive file and need to be extract in your preferred folder to install. However as this module will parse the mascot fasta files to extract sequence and description from it, it will be more efficient if installed on the same server as your Mascot Server. You should also be able to access to the database server from the server where Sequence Repository is installed.

Configuration files are located under the “<your_folder>/config”.

Datastore description

pg_uds.properties file define datastore description to access to the UDS database (for postgresql database):

javax.persistence.jdbc.driver=org.postgresql.Driver
javax.persistence.jdbc.url=jdbc:postgresql://<host>:<port>/<uds_db>
javax.persistence.jdbc.user=<user_proline>
javax.persistence.jdbc.password=<proline_user_password>

note :

• If you didn't change the default naming scheme of databases the <uds_db>= 'uds_db' so

  javax.persistence.jdbc.url=jdbc:postgresql://<host>:5432/uds_db

• proline_user_password and user_proline are the same as specified in application.conf for Proline Admin or Proline WebCore

Protein description parsing rule

As this module is used to extract Protein Sequence, description from a fasta file for a specific protein accession, it is necessary to configure the rule used to parse the protein ACC, from fasta description line. This is similar to the rules specified in Mascot Server. To do this, retrieve-service.properties file should be edited. In this file it is necessary to escape (this means prefix with '\') some characters : '\' , ':' and '='

# Name of the UDS Db configuration file (Java properties format)
fr.proline.module.seq.udsDbConfigurationFile=pg_uds.properties

# Paths must exist (regular file or directory) and multiple paths must be separated by ';' character
fr.proline.module.seq.localFASTAPaths=Y\:\\sequence;D\:\\Temp\\Admin\\FASTAs

# Java Regex with capturing group for SEDbInstance release version string (CASE_INSENSITIVE)
fr.proline.module.seq.defaultReleaseRegex=_(?:D|(?:Decoy))_(.*)\\.fasta

# UniProt style SEDb (FASTA file name must contain this Java Regex CASE_INSENSITIVE) multiple Regex separated by ';' character
fr.proline.module.seq.uniProtSEDbNames=\\AUP_;ISA_D

# Java Regex with capturing group for SEDbIdentifier value (without double quote)
# UniProt EntryName ">\\w{2}\\|[^\\|]*\\|(\\S+)"
# UniProt UniqueIdentifier (primary accession number) ">\\w{2}\\|([^\\|]+)\\|"
# GENERIC Regex ">(\\S+)"
fr.proline.module.seq.uniProtSEDbIdentifierRegex=>\\w{2}\\|[^\\|]*\\|(\\S+)

_Note_:

• fr.proline.module.seq.localFASTAPaths : only one instance should be defined. For linux system, fr.proline.module.seq.localFASTAPaths=/local/data/fasta;/local/mascot/sequence
• fr.proline.module.seq.defaultReleaseRegex : Regular expression to extract release version (CASE_INSENSITIVE) from the fasta files.
• fr.proline.module.seq.uniProtSEDbNames : Regular expression to identify Uniprot like fasta. The entry of theses files will be parse using specific rule (fr.proline.module.seq.uniProtSEDbIdentifierRegex) to extract protein accession.

For other fasta file the protein accssion will be extract by using string before first blank.

• Proline Studio application distribution is a zip file that must be unzip on each client PC. You should unzip it under a different folder than previous version

TODO