GitHub - BIMSBbioinfo/deconvR: The deconvR is an R package designed for analyzing deconvolution of the bulk sample(s) using an atlas of reference omic signature profiles and a user-selected model. (original) (raw)

deconvR : Simulation and Deconvolution of Omic Profiles

The deconvR package provides a collection of functions designed for analyzing deconvolution of the bulk sample(s) using an atlas of reference signature profiles and a user-selected model (non-negative least squares,quadratic programming, support vector regression, or robust linear regression). Users can directly use their reference atlas or, create an expended version of their reference atlas usingfindSignatures. Additionnaly, they can also use the reference atlas provided within the package, which contains cell-type specific methylation values. Another option is to simulate bulk signatures of bulk samples using simulateCellMix. And finally, we includedBSmeth2Probe function along with the Illumina Methylation EPIC B5 Manifestprobe IDs, to be used to map methylation data to respective probe IDs.

Installation

The deconvR package can be installed from Bioconductor:

if (!requireNamespace("BiocManager", quietly = TRUE)) install.packages("BiocManager")

BiocManager::install("deconvR")

You can also install the development version of deconvR directly from Github:

remotes::install_github("BIMSBbioinfo/deconvR")

How to Use deconvR

User who wish to expand their own reference atlas can usefindSignatures function. atlas is the signature matrix to be extended and samples the new data to be added to the signature matrix.atlas and samples are compliant with the function requirements. After providing appropriate atlas format, users can create samplesusing simulateCellMix function. You can get more information aboutdeconvR from here.