Capturing confounding sources of variation in DNA methylation data by spatiotemporal independent component analysis

Confounding sources of variation, which are often either unknown or known with error, are widespread in genomic datasets, and failing to adjust for them may adversely impact statistical inference. In this context, we propose a “spatiotemporal” independent component analysis method that possesses a novel invariance property, and we show that that spatiotemporal aspect may increase the ability of the method to model confounding sources of variation.