Hierarchical Canonical Correlation Analysis (HCCA)

Dataset

Dataset used in the paper can be download here.

Please make sure to unzip the Data folder to the same directory as the script files.

Package

HCCA can be run using the following command:

c = 4;
perc = 0.85;
[Us,~] = HCCA({Xexp;Xmut;Xmet;Climate}, c, perc);
X = Us{end};

Where Xexp, Xmut, Xmet and Climate are, respectively, gene expression, mutation profile, DNA methylation and Geoclimate datasets. c indicates the condition number and perc is the percentage of eigenvalues to select

Please refer to script_example_prediction.m and script_example_correlation_analysis.m for more examples of how to run the package and baselines.