DemuxHMM: Large-Scale Single-Cell Embryo Profiling via Recombination Barcoding

DemuxHMM enables recombination-aware demultiplexing of large-scale single-cell embryo scRNA-seq datasets, delivering substantial performance and scalability gains on Diversity Outbred mouse data and simulated Drosophila.

• Why it matters: Demultiplexing pooled individuals for trajectory and related analyses is constrained by limited timepoint support, either from individual-by-individual processing or accuracy loss as the number of individuals grows.
• What they did: The framework uses a Hidden Markov Model that explicitly models contiguous SNP patterns expected from simple breeding schemes, and it is tested on real Diversity Outbred mouse data and simulated Drosophila.
• The result: The recombination-aware approach improves demultiplexing performance and scales better, enabling downstream uses such as time series analysis, crossover position inference, and eQTLS in these model-organism settings.

AI summary of the abstract · may contain errors · not medical advice · read the paper.