missMethyl

missMethyl performs analysis of DNA methylation data from Illumina's Infinium HumanMethylation450 BeadChip (profiling over 450,000 CpG sites) to detect differential methylation and variability and to perform gene set analysis accounting for probe counts per gene.

Key Features:

• Normalization: Implements Subset-Quantile Within-Array Normalization (SWAN) to correct probe-type technical biases in methylation data.
• Differential Variability Testing: Uses an empirical Bayes adaptation of Levene's test to assess differential variability of CpG methylation between groups.
• Differential Methylation Analysis: Provides procedures for removal of unwanted variation to improve detection of differential methylation.
• Gene Set Analysis: Performs gene ontology-based gene set analysis that adjusts for the number of probes per gene on the array.

Scientific Applications:

• Disease Research: Identification of differential methylation patterns associated with disease to aid discovery of biomarkers and therapeutic targets.
• Developmental Studies: Analysis of methylation variability across developmental stages or conditions to study epigenetic regulation during development.
• Epigenetic Mechanisms: Investigation of mechanisms of epigenetic regulation and how environmental factors may influence gene expression via methylation changes.

Methodology:

Methods explicitly include SWAN normalization, an empirical Bayes adaptation of Levene's test for differential variability, procedures for removal of unwanted variation for differential methylation, gene ontology-based gene set analysis accounting for number of probes per gene, and implementation in R as a Bioconductor package.