ScoMAP

ScoMAP integrates scRNA-seq and scATAC-seq data into a spatial pseudotemporal virtual latent space to map enhancer-to-gene relationships and chromatin accessibility dynamics in two-dimensional tissues such as the Drosophila eye-antennal disc.

Key Features:

• Independent data generation: Generates separate scRNA-seq and scATAC-seq atlases to capture gene expression and chromatin accessibility at single-cell resolution.
• Spatial integration into virtual latent space: Embeds single-cell expression and accessibility data into a virtual latent space that reflects 2D tissue organization.
• Pseudotime ordering: Orders cells along developmental or differentiation trajectories using pseudotime to capture dynamic changes.
• Enhancer validation and mapping: Maps spatially predicted enhancers and reports that chromatin accessibility correlates with enhancer activity for approximately 85% of tested enhancers using enhancer-reporter lines.
• Inference of enhancer-to-gene relationships: Infers relationships between enhancers and their target genes within the virtual latent space, revealing regulation by multiple, often redundant enhancers.
• Deconvolution of cell type-specific effects: Uses cell type-specific enhancers to deconvolute effects of bulk-derived chromatin accessibility quantitative trait loci (QTLs) to specific cell types.
• Discovery of regulatory mechanisms: Reveals transcriptional regulatory mechanisms such as Prospero-driven neuronal differentiation via binding to a GGG motif.

Scientific Applications:

• Spatial gene regulatory mapping: Mapping gene regulatory networks and enhancer activity in spatially organized 2D tissues, exemplified by the Drosophila eye-antennal disc.
• Multimodal single-cell integration: Integrating scRNA-seq and scATAC-seq to link chromatin accessibility with gene expression at single-cell resolution.
• Developmental and differentiation studies: Characterizing dynamic gene expression and chromatin changes along developmental or differentiation pseudotime trajectories.
• Enhancer validation and functional inference: Validating spatial enhancer predictions and associating enhancer accessibility with functional reporter activity.
• Genetic variant interpretation: Deconvoluting bulk chromatin accessibility QTLs to specific cell types to interpret genetic effects on regulatory landscapes.
• Transcription factor mechanism elucidation: Identifying TF motifs and mechanisms, such as Prospero binding to a GGG motif, involved in cell fate decisions.

Methodology:

Independent scRNA-seq and scATAC-seq atlases are generated, datasets are spatially integrated into a virtual latent space, cells are ordered by pseudotime, enhancer-to-gene relationships are inferred within that latent space, and cell type-specific enhancers are used to deconvolute bulk chromatin accessibility QTLs.