Simultaneous dimension reduction and adjustment for confounding variation

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• High-throughput biological data analysis often faces challenges from confounding factors.
• Existing dimension reduction techniques may not adequately address these variations, potentially hindering accurate biological interpretation.

Purpose of the Study:

• To develop a novel method, AC-PCA, for simultaneous dimension reduction and adjustment of confounding variation.
• To demonstrate the efficacy of AC-PCA in handling biological and technical confounders in diverse high-throughput datasets.

Main Methods:

• Extension of Principal Component Analysis (PCA) to incorporate confounding adjustment (AC).
• Development of an efficient algorithm for AC-PCA with sparsity constraints for gene selection.
• Application to human brain exon array and model organism ENCODE RNA sequencing datasets.

Main Results:

• AC-PCA successfully adjusted for variations across human brain donors.
• The method effectively accounted for species-specific variations in a model organism dataset.
• Recovered anatomical structure of neocortical regions and captured shared developmental variation across species.

Conclusions:

• AC-PCA offers a robust approach for dimension reduction and confounding adjustment in high-throughput biological data.
• The method enhances the interpretability of complex biological datasets, aiding in gene selection and biological discovery.
• The developed algorithms and code are available for broader application in bioinformatics research.