MixDeR (Mixture Deconvolution in R) is a workflow (with a Shiny app) for performing mixture deconvolution of ForenSeq Kintelligence SNP data for two-person mixtures using EuroForMix and creating GEDmatch PRO reports for the individual contributor SNP profiles.
This method requires extensive validation of the settings. MixDeR provides the option of calculating various metrics for evaluating the accuracy of the deduced SNP genotypes. This is extremely useful when determining settings, specifically the allele 1 probability threshold, the allele 2 probability threshold, and the minimum number of SNPs.
Note: MixDeR (and EFM) assume the mixture samples are composed of two contributors. MixDeR is able to identify and alert the user to samples that may be potentially either single source or consist of a mixture with a large mixture ratio (i.e. > 1:100 ratio between contributors). In these scenarios, the user is warned to be cautious with the minor contributor inferred genotyping results.
The MixDeR paper can be cited with the following:
Mitchell, R., Peck, M., Gorden, E., & Just, R. (2025). MixDeR: A SNP mixture
deconvolution workflow for forensic genetic genealogy. Forensic Science
International: Genetics, 76, 103224, doi: 10.1016/j.fsigen.2025.103224
To launch the shiny app:
library(mixder)
mixder()