BioMM: Biological-informed Multi-stage Machine learning framework for phenotype prediction using omics data

Features of BioMM in a nutshell

  1. Applicability for various omics data modalities (e.g. methylome, transcriptomics, genomics).
  2. Various biological stratification strategies.
  3. Prioritizing outcome-associated functional patterns.
  4. End-to-end prediction at the individual level based on biological stratified patterns.
  5. Possibility for an extension to machine learning models of interest.
  6. Parallel computing.


BioMM installation from Github ```{r eval=FALSE} install.packages("devtools") library("devtools") install_github("transbioZI/BioMM", build_vignettes=TRUE)

BioMM has been incorporated into the [Bioconductor](
To install this package from BioConductor, start R (version "4.0") and enter:

```{r eval=FALSE}
if (!requireNamespace("BiocManager", quietly = TRUE))
BiocManager::install("BioMM", version = "4.0")


The detailed instructions on how to use this package are explained in the most updated vignette.


NIPS ML4H submission: Chen, J. and Schwarz, E., 2017. BioMM: Biologically-informed Multi-stage Machine learning for identification of epigenetic fingerprints. arXiv preprint arXiv:1712.00336.

Chen, Junfang, et al. "Association of a Reproducible Epigenetic Risk Profile for Schizophrenia With Brain Methylation and Function." JAMA psychiatry (2020).

Try the BioMM package in your browser

Any scripts or data that you put into this service are public.

BioMM documentation built on Nov. 8, 2020, 11:04 p.m.