Cell composition correction software tools | DNA methylation microarray data analysis
Whole blood is frequently utilized in genome-wide association studies of DNA methylation patterns in relation to environmental exposures or clinical outcomes. However, for DNA methylation assessed from whole blood, the association between DNA methylation and an exposure of interest could be confounded by cellular heterogeneity. In larger epidemiological studies, it is not feasible to isolate and profile every individual cell subset. Thus, several algorithms have been developed to measure and adjust for cellular heterogeneity in whole blood.
A suite of computational tools that incorporate state-of-the-art statistical techniques for the analysis of DNAm data. minfi provides methods for preprocessing, quality assessment and detection of differentially methylated regions from the kilobase to the megabase scale. Several preprocessing algorithms are available and the infrastructure provides a convenient way for developers to easily implement their techniques as Bioconductor tools. By making SNP annotation available, users can choose to be cautious about probes that may behave unexpectedly due to the inclusion of a SNP in the probe sequence. minfi is unique in that it provides both bump hunting and block finding capabilities, and the assessment of statistical significance for the identified regions. Finally, because the package is implemented in Bioconductor, it gives users access to the countless analysis and visualization tools available in R.
An R package for comprehensive analysis of DNA methylation data obtained with any experimental protocol that provides single-CpG resolution, including Infinium 450K microarray and bisulfite sequencing protocols, but also MeDIP-seq and MBD-seq.
Allows Illumina HumanMethylation BeadChip analysis. ChAMP is an integrated analysis pipeline including functions for (i) filtering low quality probes, adjustment for Infinium I and Infinium II probe design, (ii) batch effect correction, detecting differentially methylated positions (DMPs), (iii) finding differentially methylated regions (DMRs) and (iv) detection of copy number aberrations. The software also allows detection of differentially methylated genomic blocks (DMB) and Gene Set Enrichment Analysis (GSEA).
A method based on principal component analysis (PCA) and designed for the correction of cell type heterogeneity in epigenome-wide association studies (EWAS). ReFACTor tool is based on a variant of PCA and can be applied to any tissue. It selects the sites that can be reconstructed with low error using a low-rank approximation of the original methylation matrix. Moreover, ReFACTor does not use the phenotype in the selection process, making ReFACTor useful as part of a quality control step in EWAS.
Permits reference-free deconvolution. RefFreeEWAS offers a method for evaluating the extent to which the underlying reflects specific types of cells. It differs from widely used principal components analysis (PCA)-based methods such as Surrogate Variable Analysis (SVA) in imposing biologically based constraints, thus resulting in mixture coefficients having greater biological interpretation and placing greater emphasis on coordinated cellular processes.
An R version of FaST-LMM-EWASher, which performs epigenome-wide association analysis in the presence of confounders such as cell-type heterogeneity. A python version of this software is also available as part of Fast-LMM-Py.