Contents - Index

Data Manager

The data sets defined in the Control panel are managed with the Data manager. It is open by either pressing the Data manager button, or choosing Data Manager in the Tools menu, both found in the main window. The settings made in the Data manager apply to all data files in a data set. If a data set is cloned, the new clone will include the same data files as the original, as well as all settings made in the Data manager. The settings of the original data set and the clone can then be adjusted separately.

There are three tabs in the Data manager: Data selection, Colors & Symbols, and Groups, which includes the following options.

Data selection: 

• Activate/inactivate samples for analysis. Only the active samples are used in the plotting and analyses.
• Activate/inactivate genes for analysis. Only the active genes are used in the plotting and analyses.
• Set attribute (training/test) to samples. Models built in with classification methods are based on the training samples and tested on the test samples. 
• Set attribute (predictor/response) to genes and variables. Defines which columns are used as input to a model (predictors), and which columns that represent the output (response).
• Mean center the data reversibly.
• Autoscale the data reversibly.

Colors & Symbols:

• Assign colors and plot symbols for plotting of samples or genes.
•  
• Groups:
• Define groups of rows. This is used e.g. in some statistical methods such as t-test, where you test if there is a differance in group means.

In the top middle of the main window you find the Transpose button that twists the data in such a way that columns becomes rows and vice versa. If you transpose your data set, any scaling you have applied in the Data manger will be kept as it were. Scaling e.g. columns before you transpose the data set will therefore result in scaled rows in the transposed data. Since groups and coloring are always defined for rows, any grouping or coloring made before transposing a data set will not be active in the transposed data. If you transpose the data back to its original form (press the Transpose button once more), the groupings and coloring will become active again. It might be a good idea to analyze the data in both dimensions. This is done by first managing the data in its original form and performing the analysis, after that you transpose the data and repeat the managing and analysis.  

Watch the Scatter plot tutorial on qPCR based NHL diagnostics on www.multid.se. Here you can see how samples are grouped together and colored according to their groups.