Disambiguate

Disambiguate separates sequencing reads from two species in grafted samples by operating on DNA and RNA-seq alignments to enable accurate variant calling and gene expression quantification in mixed-species (e.g., human-mouse) datasets.

Key Features:

• Species Separation: Processes DNA and RNA-seq alignments to distinguish reads originating from two species, including human-mouse mixtures, with reported high sensitivity and specificity.
• Recovery of Tumor Signal: Separates mixed-species components to maximize recovery of target tumor reads, improving downstream variant calling and gene expression quantification.
• Alignment-based Processing: Operates directly on sequencing alignments to assign reads to species-level origin.
• Implementations: Provided in both Python and C++ implementations for integration into computational pipelines.

Scientific Applications:

• Grafted-sample analysis: Enables separation of host and graft reads in xenograft and other grafted-sample experiments for accurate genomic and transcriptomic analysis.
• Preclinical drug-development studies: Supports improved variant detection and expression profiling in preclinical models that combine human tumor and non-human stroma (e.g., human-mouse xenografts).

Methodology:

Processes DNA and RNA-seq alignments from two species to assign reads to species of origin; implementations are available in Python and C++.