The correlation analysis can be used to compare patterns among matrix columns or rows, for example to compare patterns of expression among genes for different samples.
Before running a correlation analysis you should have this data prepared in files:
A matrix with the data resulting of your experiment (i.e. an expression matrix) or a list or lists of elements (i.e. genes or probesets) that has been found significant after other analysis (i.e. after differential expression you get lists of genes over and under expressed).
To start the wizard go to the menu File > New > Analysis > Correlation analysis ...
You can also start a correlation analysis from an opened heatmap by selecting the menu Analysis > Correlation.
scale: | 50% :align: center
|
---|
You can avoid this first presentation page by selecting the option Don’t show this page next time.
scale: | 50% :align: center
|
---|
scale: | 50% :align: center |
---|
Transform to 1 / 0: Some statistical tests are designed to work with discrete events (as Binomial or Fisher’s exact tests), this option allows to transform a matrix with real values into a binary matrix containing only 1’s and 0’s for the analysis. All the values which satisfy the condition will be transformed to 1 and the rest to 0. For example, if the data file is an expression matrix with log 2 ratios it can be transformed to a binary matrix having a 1 for all the log 2 ratios greater than 1.5. Other possible application is when the matrix have p-values, if a significance of 0.05 is considered, all values less than 0.05 could be transformed to 1’s.
scale: | 50% :align: center
|
---|
The option Apply to allows to select whether to correlate columns or rows.
scale: | 50% :align: center
|
---|
scale: | 50% :align: center
|
---|
After performing the analysis a new editor with the details of the analysis will be opened:
To explore the data matrix resulting from the transformations with a heatmap click on the Heatmap button in the Data section.
To explore the results of the analysis with a heatmap click on the Heatmap button in the Results section.
scale: | 50% :align: center |
---|