SpaMAP: robust integration and annotation of single-cell and spatial omics data using interpretable deep neural networks

Spatial Mapping of Single-Cell Transcriptomic Data This Python script performs spatial mapping of single-cell transcriptomic data using a two-stage strategy: step1. Mutual Nearest Neighbor (MNN) Matching: Maps a subset of single cells to spatial sequencing spots, generating high-confidence spatial anchors. step2. Domain-Adversarial Neural Network (DANN / GRNN): Learns domain-invariant representations between mapped and unmapped single cells. The model predicts anchor spatial positions for mapped cells and aligns unmapped cells in a shared latent space, enabling spatial position inference for previously unmapped cells. Example Usage: python SpaMAP.py spatial_profile scRNA_profile