PCA-Plus: Enhanced principal component analysis with illustrative applications to batch effects and their...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Principal Component Analysis (PCA) is standard for analyzing large biomedical datasets.
• Conventional PCA struggles with moderate batch/trend effects in multi-omic data (e.g., TCGA).
• Objective quantification of batch effects is challenging with standard PCA.

Conclusions:

• PCA-Plus is a valuable tool for analyzing, visualizing, and quantifying effects in molecular profiling data.
• The software has been successfully applied in numerous NCI-funded projects (TCGA, PanCancer Atlas).
• PCA-Plus algorithms are generic and available as a free R package.