scMET
scMET models single-cell DNA methylation using a hierarchical Bayesian framework to quantify methylation levels and biological heterogeneity while addressing sparse CpG coverage.
Key Features:
- Hierarchical Bayesian model: employs a hierarchical beta-binomial specification integrated within a generalized linear model to capture biological overdispersion.
- Information sharing across cells and genomic features: leverages pooling of information across cells and genomic features to mitigate sparse CpG coverage.
- Quantification of variability: provides biologically interpretable estimates of methylation variability and identifies highly variable features.
- Differential analysis: performs differential methylation and differential variability analyses between pre-specified groups of cells.
- Feature selection: implements feature selection to characterize epigenetically distinct cell populations and identify key regulatory elements.
Scientific Applications:
- Characterization of epigenetic heterogeneity: study DNA methylation heterogeneity at single-cell resolution to inform gene regulation.
- Formulation of biological hypotheses: estimate methylation variability to generate hypotheses about epigenetic regulation, including during early developmental stages.
- Analysis of large-scale datasets: applicable to recent large-scale single-cell methylation datasets for comparative and population-level analyses.
Methodology:
Uses a hierarchical Bayesian framework with a beta-binomial model embedded in a generalized linear model, incorporates information sharing across cells and genomic features, and supports inference on overdispersion as well as differential methylation and differential variability.
Topics
Details
- License:
- GPL-3.0
- Programming Languages:
- R, C++
- Added:
- 1/18/2021
- Last Updated:
- 2/13/2021
Operations
Publications
Kapourani C, Argelaguet R, Sanguinetti G, Vallejos CA. scMET: Bayesian modelling of DNA methylation heterogeneity at single-cell resolution. Unknown Journal. 2020. doi:10.1101/2020.07.10.196816.