LabSystemAdmin

Each installation of the lab system needs some administration. This gives an overview about what needs to be done:

Pre-installation

The system is independent from hardware types, OS used and (in theory) also from the type of NMR machines used. You need a machine for the nmrshiftdb2 system to run on. It probably makes sense to use a dedicated server, although this is not a strict requirement. Hardware requirements are modest, any current CPU with standard amount of memory should work fine. Space on harddisk is also not demanding. One lab system reached, together with the data from the public nmrshiftdb2 instance, around 1.5 GB after several years of operation. More space might be needed for the raw data (see below).

The lab system must have a directory (configured in nmrshiftdb.properties file) on the server it runs on which contains a sub-directory for each machine (the "raw file directory"). The system assumes that for each order a directory with the name of the order is created in the directory of the respective machine. E. g. an order named "myorder" has been processed on the machine named "av600", the system assumes that a directory /path/from/nmrshiftdb.properties/av600/myorder is created. In our current installations, this is done by pushing the files on the server using Samba via a script triggered by the Bruker software. Other configurations would be possible, e. g. the directory could be somewhere else and nfs-mounted on the nmrshiftdb2 server, as long as the software sees the directory.

nmrshiftdb2 packs the data for each order into a zip and places this in a directory. It also creates a link to this file where the file name is appended to a configured first part. So you need a place which can be addressed a) as a local directory for the software (again the directory might actually be mounted) and b) via an url (might be any protocoll).

The following data are needed during the installation and it is probably best to explore these before:

• Which machines exist? For each machine: NAME, FIELD_STRENGTH (1H), allowed for self/operator/changer operation, which of these should have compulsory structure input?
• Which experiments are possible? For each experiment: Name, should it appear as checkbox/dropdown box in the order form, if dropdown box, what is the maximum number?
• How many columns should be in the order form? Which experiment goes where?
• Which categories should exist in reports (e. g. 1D, 2D)? Which experiments belong where?
• Which experiments are allowed in which machine?

Installation

For installation follow the admin guide in the doc directory of the nmrshiftdb2 checkout. Note the additional properties in the nmrshiftdb.properties file. For a lab system, in addition the following tasks are needed:

• Populate the tables LABGROUP, MACHINE, EXPERIMENT_CATEGORY, EXPERIMENT_REPORT_CATEGORY, EXPERIMENT_MACHINE, EXPERIMENT as needed
• Deploy nmrshiftdb2 as described in the admin guide
• Make sure the raw file directory is set up properly
• Run the script populate-labsystem from src/sql of the nmrshiftdb2 checkout (this creates a user nmruser as labgroup leader and two users testuser and testworker. The user nmruser can be changed and must not be the labgroup leader, whereas testuser and testworker are needed for technical reasons and must not be removed or altered. They can also be used for playing around, since orders from these users are not included in the statistics).
• From now on, further users can be added as described below.

Post-Installation

The labgroup and the labgroup leader

A "labgroup" is a set of spectrometers and associated users. Typically, one labgroup exists for one lab and runs on one installation. It is possible to create more than one labgroup for a lab, this could for example mean a separate administration of a set of spectrometers.

Each labgroup has a "labgroup leader". This is an nmrshiftdb2 user who can administer machines, experiments etc. There can be more lab workers, who run experiments and can administer users. The labgroup leader must exist in the database (typically created by normal registration). The ID of the labgroup leader is entered in the table LABGROUP (needs to be done manually) and the user needs the labgroup_leader role assigned (needs to be done manually). Also the PSML profile of nmruser must be used by this user.

The spectrometers

If the labgroup leader logs in to nmrshiftdb2, the "NMR lab administration" tab should be active. The following tasks need to be done here;

• Add spectrometers: Add spectrometers using the "Add a new one"-box below the list of spectrometers. The frequency to enter is the 1H frequency, typically part of the spectrometer name. Once added, the spectrometer can be edited in the list. The following parameters can be changed:
  • There are four possible modes, Sample Changer/Barcode/Operator Order/Self Order. A working day is divided in two shifts, for each of them (called "day" and "night") modes can be activated. If it is more than one, the user can chose. If "Structure input compulsory?" is set for a mode, a structure is needed for submissions in this mode.
  • The shift times are set in the "Shift operation" column. The first time is for changing from night to day (typically start of the working day), the second from day to night (typically late afternoon). Leaving "Shift operation" off, results in night mode for the whole day.
  • "Default Mode Day/Default Mode Night" is the mode which is displayed with a new order in the respective shift. The actual mode can always be chosen by the user.
  • "Time for sample change" is used for accounting. The time is added to the time from the bruker audita.txt files.
  • If "IconNMR integration?" is on, an automation file will be written to the directory path. This file is transferred using scp, with the data set in the interface.
  • A spectrometer can be deleted by checking both boxes in "Check twice and click to delete this machine" and hitting change. If one box is checked (accidentially), nothing happens (note that "deleting" only hides a machine, it is kept to keep track of historical activity).
  • It is possible to make a spectrometer to be displayed on top of the list on a specific computer (typically the computer controlling that spectrometer). For this, the IP needs to be set manually in the database (field DEFAULT_ON_MACHINE in table MACHINE).
• Add experiment: The next box shows the experiment which can be executed. This is independent of spectrometer, the enabling on specific spectrometers follows later. New experiments are created using the "Add a new one"-box below the list of experiments. Only the name is needed here. Once created, the experiments can be changed in the list. The following settings can be changed:
  • Max. Number: TODO.
  • Subexperiments (if composite): An experiment can be composed of several other experiments. None means the experiment is not a composed one.
  • Identical with: TODO
  • Report category: Indicates which category this experiment is added in the reports. Only influences the report. Categories need to be created manually in the database.
  • Form tab: The tab this experiment shows up on the submission page. Tabs need to be created manually in the database.
  • Text before: This is any text displayed before this experiment. It can also be a html tag like a line break to structure the experiment list.
  • Automation order: These numbers determine the order the experiments are written in the automation file. If two or more experiments have the same number, their order is random.
  • Order on screen: These numbers determine the order the experiments are shown on screen for users. If two or more experiments have the same number, their order is random.
  • An experiment can be deleted by checking both boxes in "Check twice and click to delete this machine" and hitting change. If one box is checked (accidentially), nothing happens (note that "deleting" only hides an experiment, it is kept to keep track of historical activity)
• Add experiments to machines: In the next box, experiments can be activated for individual machines. This is done separately for day and night shift. If shift mode is not used, the night values apply. The "Name in IconNMR files" is the name which is written to the automation file if automation is used. This can be different for each spectrometer and must match with the parameterfile of TopSpin. The duration will be added to the duration from the audita.txt file, can be set to 0 if not relevant. "Compulsory" means that the experiment will be included in every order on this machine and cannot be deselected.
• Solvents: For automation mode, every spectrometer has an individual solvent list. Solvents can be added from the list or by setting an individual name followed by pressing “Add solvent”. After adding all solvents the order of the solvent list can be changed individual for every spectrometer by changing the order number behind the solvent name. To delete a solvent, check both boxes and press “Remove/order solvents”. Only change the order for one spectrometer at ones. For operator and self measurements, there is one list only (and the users can enter any value they like). This list is maintained directly in the database.

Lab workers

Lab workers can finish order for worker mode. They can also do user administration (as well as the lab leader, whereas the setup above is restricted to the lab leader). Lab workers can register and chose the "I am an operator in this lab group" box. They need to get approved by the lab leader. Alternatively, the role labgroup_worker can be set manually in the database.

User administration

Users register themselved via the normal registration process (make sure users use the lab server for this, not the www instance of nmrshiftdb2). They need to check the "I want to get my compounds measured there" checkbox and chose their labgroup. All labworkers can see a list of users. The new users need to be activated (Email notifications are provided). The lab needs to decide on a policy here.

MNova licenses can be administered via the web interface.