diem: Diagnostic index expectation maximisation

NEWS 30th October 2025: diem development continues in R and Python with a new fully Pythonic bedtools-aware implementation at this permanent institutional site https://github.com/Studenecivb/diemPy

This site (here, personal to SJEB) will remain as a record of the original implementations and to serve as a re-direct for Python users (above) and R users (below).

<img src="/diemimages/CircleDiagram24MayBitMap.jpg" alt="Alt text" title="" style="display: inline-block; margin: 0 auto; max-width: 300px">

diem is a genome polarisation algorithm that finds which alleles of genomic markers belong to which side of a barrier, co-estimating which individuals belong on either side. diem is a deterministic expectation maximisation algorithm and so uses the likelihood inference framework.

The diem algorithm has been implemented in several programming languages, with core functions and accompanying tools stored in this repository.

1. diemr - R implementation. The package developer version is available through https://github.com/nmartinkova/diemr and the stable release through https://CRAN.R-project.org/package=diemr.
2. diempy - Python implementation.
3. diemmca - Mathematica implementation. Soon also to be available at https://www.notebookarchive.org

When using any of the diem algorithm implementations, cite:

Stuart J. E. Baird, Jan Petružela, Izar Jaroň, Pavel Škrabánek, Natália Martínková. 2022. Genome polarisation for detecting barriers to geneflow. bioRxiv 2022.03.24.485605; doi: https://doi.org/10.1101/2022.03.24.485605.

Stuart J. E. Baird, Jan Petružela, Izar Jaroň, Pavel Škrabánek, Natália Martínková. 2023. Genome polarisation for detecting barriers to geneflow. Methods in Ecology and Evolution; doi: https://doi.org/10.1111/2041-210X.14010.