Phase 3: Correlation analysis

Purpose

Correlation analysis helps understand the relationships between different immune measurements across all samples and timepoints. This helps confirm if certain responses tend to occur together (positive correlation) or are mutually exclusive/inversely related (negative correlation)

Action

Perform correlation

1. Navigate to perfrom correlation analysis by going to Discovery -> Start -> Correlation

1. Choose the same numerical immunological assays columns as used in PCA under Column Selection

2. Select Spearman for the Correlation Method within the Column Selection Tab

1. Under the Preprocessing tab, select center and scale to normalize the data

1. Go to the Correlation Settings tab:

   1. NA Action: Set it to a method that can appropriately handle missing values such as pairwise.complete.obs

   2. Plot Type: Select preferred option to view the correlation in the plot. For this example, the Full type was chosen

   3. Reorder Correlation: Select Hierarchical clustering to visualize relationships between clustered parameters

   4. Method: This tab will appear when Hierarchical clustering is selected. Select Ward algorithm for clustering.

Correlogram - Correlation of Immune parameters

Correlogram Analysis: Correlation of Immune parameters and Timepoints