Vireo

Vireo demultiplexes pooled single-cell RNA-seq (scRNA-seq) data using a Bayesian model to assign cells to donor samples for downstream gene expression analysis.

Key Features:

• Bayesian Framework: Employs a Bayesian model to probabilistically assign cells to samples within pooled scRNA-seq datasets.
• Genotype Independence: Performs demultiplexing without requiring complete genotype information from the pooled samples.
• Natural Genetic Variant Barcoding: Utilizes natural genetic variants as endogenous barcodes to distinguish cells originating from different donors.
• Computational Efficiency: Designed for computational efficiency to scale to large single-cell RNA-seq datasets.

Scientific Applications:

• Multiplexed scRNA-seq demultiplexing: Recovers sample identities in pooled experimental designs to enable analysis of multiple samples from a single run.
• Comparative gene expression studies: Facilitates comparison of gene expression across conditions or time points while mitigating batch effects from separate library preparations.

Methodology:

Vireo applies Bayesian inference to genetic-variant signals to reconstruct sample identities without requiring complete genotype data and was validated on synthetic mixtures and real-world scRNA-seq datasets.