Working with datasets
The dataset view is the 'home page' for the uploaded dataset. It provides information about the dataset, allows you to run analyses and gives the status of submitted jobs (figure 1). The sections below give more details about the available functionality on this page.
Figure 1: The dataset page
[Top]Accessing a dataset
To access an uploaded dataset, unregistered (anonymous) users will need to visit the web address of the dataset's page, which is can be saved (e.g. bookmarked) during the upload process. Registered users have a list of their datasets on the My datasets page, and can access a particular dataset by clicking on its name.
[Top]'About this dataset' panel
Figure 2: The About this dataset panel
The panel on the right hand side of the dataset page gives information about the dataset, allows editing of the dataset's notes and enables attachment of survival data (figure 2), and downloading of the TMA data and survival data, if previously uploaded. The information listed in the panel is as follows:
- Name - The name of the dataset, as entered when the dataset was uploaded.
- Notes - Notes on the dataset, optionally entered when the dataset was uploaded. The notes are editable (see editing the dataset's notes).
- Date uploaded - The date and time (GMT) the dataset was uploaded to TMA Navigator.
- Available until (anonymous users only) - The date and time (GMT) that the dataset will be deleted from TMA Navigator. This is six months from the date of upload.
- Survival data - Indicates whether the dataset has survival data attached or not - indicated by Not present or Present. If survival data is not present, a button is provided to attach the data (see attaching survival data); if present, the number of samples with attached patient survival data is shown.
- Scoring - Indicates the type of scores in the dataset - either continuous or categorical. Usually, this is automatically selected during upload. Continuous scoring systems provide a continuous range of scores (e.g. AQUA). Categorical scoring systems have a fixed number of categories (e.g. Allred). The scoring type determines which analyses are available.
- Samples - Indicates the number of samples (e.g. patients) in the dataset.
- Markers - Indicates the number of markers (typically proteins) in the dataset.
- Batch correction (if applicable) - If the dataset has multiple batches, the batch correction method applied will be displayed here.
- Replicate aggregation (if applicable) - If the dataset has multiple replicates, the aggregation method applied will be displayed here - either Mean or Median.
- Download data - the data uploaded to TMA Navigator can be downloaded including any processing applied during import. Click the TMA button to download the marker scores in TSV format. If survival data has been uploaded, it can be downloaded in TSV format by clicking on the Survival button.
Editing the dataset's notes
To edit the notes on a dataset, click directly on the notes, or on the message 'Click here to edit'. A text box appears in the notes area allowing editing. When complete, click on the save button. Clicking on the cancel button, or elsewhere on the page, will discard any changes.
Attaching survival data
Figure 3: The dataset page with the Attach survival button highlighted.
Figure 4: Dialog for attaching survival data to a TMA dataset
Survival data is required to perform certain types of analyses - Kaplan-Meier plots of marker scores against survival, for example. Please refer to formatting requirements for the survival file to prepare the data.
Once you have a file ready for upload, go the dataset view for the relevant dataset, and click the Attach survival button. The button is located in the About this dataset panel on the right-hand side (figure 3). A dialog box will then appear (figure 4). Click on Upload a file and select the survival file on your computer. After a few moments, the survival file will be uploaded and subject to validation. If the file validates successfully, a confirmation box will appear stating that survival data has been attached. If the file does not validate, an error message will be displayed, and there will be the opportunity to upload a new file after correcting any issues raised in the error message.
[Top]Running analyses
Figure 5: The Run analysis dialog with the Mixture model plots analysis selected
To choose an analysis, click the Run analysis button at the top left of the dataset page. A dialog box will then appear allowing you to choose an analysis (figure 5). On the left is a list of available analyses, the right side gives information about the selected analysis. The list of analysis categories shows the name of each category, followed by the number of available analyses in brackets. Click on an analysis category, for example Data exploration, and a list of analyses in that category will appear below. Selecting an analysis from the list gives a summary of the analysis on the right-hand side. To select an alternative analysis, use the list on the left hand side. For more information on the selected analysis, click the Help button on the top right.
To start the selected analysis, click the Run analysis button in the dialog box (Figure 5). After a few moments, the box will close and the job will be added to the Pending analyses table at the bottom of the page. Once the analysis has finished, it will move from the Pending analyses table to the Completed analyses table.
To view the results of an analysis, click on its name in the Completed analyses table.
If an error occurs running an analysis, the message Error, with a Help button next to it, will appear under Completion date. Click on the button for more information on why the analysis was unable to run, which is often due to the specifics of the dataset.
[Top]Pending analyses
Figure 6: The 'Pending analyses' table, which shows analyses currently being processed or in a queue
Pending analyses are shown in a table (figure 6) beneath any completed analyses. If there are no analyses currently queued or processing, a message is displayed to that effect.
Pending analyses are listed in the table with the following information:
- Name - A name to describe the analysis type.
- Submission date - The date and time (GMT) the analysis was submitted (when it was requested).
- Status - The status of the job, either queued or processing. When the analysis is complete, it is automatically moved to the completed analyses table.
- Queued - The analysis is in a queue waiting to run, and so the the results are not ready yet.
- Processing - The analysis is currently being processed and the results should be ready soon. For compute-intensive analyses, the progress is displayed as a percentage.
[Top]Completed analyses
Figure 7: The 'Completed analyses' table
The completed analyses table (figure 7) shows a list of analyses that have been processed, grouped by category. To view the results of a successful analysis, click on its name. For more information on interpreting results, see analysis types.
The table contains the following information:
- Name - A name to describe of the analysis type.
- Completion date - The date and time (GMT) that the analysis was completed. If an error occurs when running an analysis, the message Error will appear, with a Help button next to it. More information on why the analysis did not complete successfully is available by clicking on the Help button. Also see the error messages section of the help for further details.