LifangHe/sCCA

Sparse Canonical Correlation Analysis (sCCA)

Integrating multi-OMICS data through sparse Canonical Correlation Analysis for the prediction of complex traits: A comparison study

This repository presents the code of the methods, simulation studies and data analyses performed in "Integrating multi-OMICS data through sparse canonical correlation analysis for the prediction of complex traits: a comparison study" by T. Rodosthenous, V. Shahrezaei and M. Evangelou, published in Bioinformatics, Volume 36, Issue 17, Pages 4616–4625.

(A) sCCA Functions:

• ConvCCA on LASSO and SCAD penalties, for two or more input datasets
• RelPMDCCA on LASSO and SCAD penalties, for two or more input datasets
• Function in obtaining additional canonical pairs

(B) Simulation Analysis

• Data generating models
• Performance assessment

(C) Data Analysis

• nutriMouse
  • Data are found under the R package "whitening", by running data("nutrimouse")
• cancerTypes
  • Data are found in the following link: http://compbio.cs.toronto.edu/SNF/SNF/Software.html