novoSpaRc

novoSpaRc reconstructs spatial locations of single cells within tissues from single-cell RNA sequencing (scRNA-seq) data to enable study of tissue organization and spatial gene expression patterns.

Key Features:

• Probabilistic Spatial Assignment: Employs a probabilistic approach based on the structural correspondence hypothesis that cells in close physical proximity exhibit similar gene expression profiles.
• Reference Database Utilization: Does not require an existing marker-gene reference atlas but can optionally incorporate a reference database of marker genes to improve reconstruction performance and accuracy.
• Efficient Processing: Capable of mapping scRNA-seq datasets of 10,000 cells onto 1,000 tissue locations in under five minutes.

Scientific Applications:

• Mouse Organ of Corti Reconstruction: De novo reconstructed the mouse organ of Corti using the structural correspondence assumption to model complex tissue structure.
• Human Osteosarcoma Cell Cultures: Mapped human osteosarcoma cultured cells by incorporating marker gene information to resolve spatial locations.
• Drosophila Embryo Reconstruction: Demonstrated reconstruction of Drosophila embryos combining structural correspondence with marker-gene strategies, with a step-by-step protocol provided.

Methodology:

Applies the structural correspondence hypothesis to analyze scRNA-seq gene expression profiles and probabilistically assign cells to positions within a tissue matrix, with optional incorporation of a reference atlas of marker genes.