Guide to Accessing Neuroimaging Analysis Network (NaN)

Prerequisites

Before accessing the NaN platform, ensure you have the following:

1. VPN Connection to KCL Intranet: As a first step you need to connect to King's intranet. The links below are not accessible outside of KCL's network perimeter.
2. Neuroimaging Analysis Network (NaN) Account: Create an account at NaN for accessing various neuroimaging tools and resources. For account creation and more information, visit NaN Accounts.
3. Remote Desktop with NoMachine: Install and configure NoMachine for remote desktop access. Learn more about setup and usage at NoMachine Guide.

File Sharing and Data Transfer

To perform the analysis, it's essential to copy the applications and the data for analysis to the remote computer. Cyberduck is a recommended tool for this purpose. You can download it from Cyberduck's official website.

Setup connection to NaN server

1. Open a New Connection: Click on Open new connection.

2. Configure the Connection:

   • Choose SFTP (SSH File Transfer Protocol) from the dropdown.
   • Enter your credentials (username and password) for the remote server.

   

3. Access Remote Desktop

• Once connected, a new disk labeled desktop.nan.kcl.ac.uk will appear.
• Click on this disk to access the file system of your remote desktop.

Transfer Files

1. You can now drag and drop files from your local computer to the remote desktop.
2. The files you need to transfer include:

• MATLAB Application Files: Upload the provided MATLAB file. This file must be added to Matlab's path on the remote computer for the applications to work (see run the applications).
• Data for Analysis: Upload these files to a separate folder designated for data.

By following these steps, you'll be able to efficiently transfer necessary files and data to your remote desktop for analysis.

Start Matlab on Remote Desktop

After establishing a connection to your remote desktop, follow these steps to start MATLAB. This process involves a few extra steps beyond the usual method.

1. **Connect a processing machine via Nan Terminals **

   1. Navigate to Applications -> Nan Terminals.
   2. Choose a terminal from the available options (see figure: left hand side). This selection connects you to a machine powerful enough to run your analysis.
   3. Your NaN desktop acts merely as an entry point to these instances. Once you select a terminal, you interact with it through a terminal window (see figure: central window).

2. Launch MATLAB:

• In the shell window of the terminal, type commands to start MATLAB running on the selected machine.
• Enter the following commands: \

     >>> module load matlab/latestmatlab
     >>> matlab
  

• Wait for MATLAB to start.

Run the Applications

To run the applications on MATLAB, follow these steps:

• Add MATLAB Folder to Path: Add the MATLAB folder, which contains the applications and was previously copied to the remote desktop, to MATLAB's path.

  

• Start the Applications: Simply type Gazelle or Coconi in MATLAB's command window to launch the applications.

❗️If the application windows are not within the visible area of your desktop, you can reposition them. Right-click on the MATLAB window, hover over the window of the application you want to move (e.g., Gazelle or Coconi), and select move to bring it to a convenient area on the desktop.