# MAGAR
[](https://travis-ci.org/MPIIComputationalEpigenetics/MAGAR)
[](https://www.gnu.org/licenses/gpl-3.0)
## Short description
*MAGAR* (Methylation-Aware Genotype Association in R) is an R package for computing DNA methylation quantitative trait loci (methQTL) from DNA methylation and matched genotyping data. The tool employs a two-stage approach to compute methQTL:
- CpG correlation blocks are computed to mimic DNA methylation haplotypes. The correlation blocks are computed from CpGs that are highly correlated across the samples.
- For each correlation block, a representative CpG (tag-CpG) is computed and used in a linear modeling strategy (or alternatively with [*fastQTL*](http://fastqtl.sourceforge.net/)) to determine significant associations between genotype variation and changes in DNA methylation of particular CpGs.
The package furthermore uses established software tools for handling DNA methylation and genotyping data, including [*RnBeads*](https://rnbeads.org), [*PLINK*](http://zzz.bwh.harvard.edu/plink/), and [*CRLMM*](https://www.bioconductor.org/packages/release/bioc/html/crlmm.html). Thus, raw DNA methylation data (IDAT, BED files) and genotyping data (IDAT, PLINK files) can be used as input.
## Installation
The package needs some Bioconductor dependencies, most notably *RnBeads*, which can be directly installed from the [*RnBeads* website](https://rnbeads.org). Atfer instaling *RnBeads*, *MAGAR* can be installed in a R session using the ```devtools``` package
```r
source("https://rnbeads.org/data/install.R")
if(!requireNamespace("devtools")) install.packages("devtools")
devtools::install_github("MPIIComputationalEpigenetics/MAGAR")
```
## Documentation
You can find a detailed information for *MAGAR* [here](vignettes/MAGAR.Rmd)
## Citation
Scherer, M., Gasparoni, G., Rahmouni, S. et al. Identification of tissue-specific and common methylation quantitative trait loci in healthy individuals using MAGAR. Epigenetics & Chromatin 14, 44 (2021). [doi:10.1186/s13072-021-00415-6](https://doi.org/10.1186/s13072-021-00415-6)
## Contact
You can contact [Michael Scherer](mailto:michael.scherer@crg.eu) for reporting bugs, feature requests or questions.
