A Two-Pronged Strategy Eliminates Dissociation Artifacts for High-Fidelity Neuroimmune Single-Cell Transcriptomics

Here, we present a comprehensive two-pronged strategy to eliminate these artifacts in key neuroimmune niches. First, we established a benchmark by applying an optimized experimental protocol that minimizes dissociation stress.

Comparative analysis against conventional methods allowed us to map the landscape of SAT artifacts across cell types in the brain and skull bone marrow (SBM), defining distinct, tissue-specific signatures. We then employed random forest-based approach to refine these signatures into powerful SAT gene panels, creating a computational tool for the sensitive detection and removal of artifactual signals. Our experimental approach achieves a ~95% reduction in these artifacts, while computational approach alone also has a 60-64% reduction that enables the retrospective correction of published data.