Welcome to 3d_OT’s documentation!

Contents:

• Installation
  • System Requirements
• Installation via Github
  • Download
  • Environment
  • Package
• Jupyter Tutorial
• Single Omics Spatial Domain Identification
  • Spatial Clustering Demonstration of Breast cancer (10xVisium)
  • Spatial Clustering Demonstration of 12 Slices from the DLPFC Dataset(10xVisium)
  • Spatial Clustering Demonstration of mouse visual cortex(Starmap)
• Multi Omics spatial domain identification
  • Spatial multi-omics Clustering Demonstration of five simulated data
  • Spatial multi-Omics Clustering Demonstration of Human_Lymph_Node A1
  • Data integration for mouse brain Spatial-epigenome-transcriptome
• Single Omics alignment
  • Alignment tutorial for 8months mouse brains
  • Alignment tutorial for 151673 and 151674 from DLPFC
  • Alignment tutorial for two E15.5 Stereo-seq mouse embryo slices
• Mouse brain p(22) alignment
  • Single-Omics alignment of mouse brain of H3K27ac+RNA with ATAC+RNAmouse brain
  • Alignment tutorial for Multimodal alignment
• 3D reconstruction
  • Loading package
  • loading data
  • We visualize every dataset in 2D before alignment
  • Building spatiotemporal developmental trajectory using 3d-OT
  • Visualize the temporal developmental trajectory of the Heart
  • Visualize the temporal developmental trajectory of the Liver
  • Get spatiotemporal developmental trajectory files and sankey plot
• Cross_platform
  • Alignment tutorial for seq-FISH and Stereo-seq
  • Loading package
  • 1. Loading and Pre-processing two slices from E8.75seq-FISH mouse embryo and E9.5 Stereo-seq mouse embryo
  • 2. Constructing neighbor graph and training the Pointnet++Encoder

Overview

3d-OT leverages the PointNet++ model and optimal transport with soft communication as its foundation. Through a modular integration approach, it seamlessly supports single-modal, multi-modal, and diverse spatial omics data from various technologies. By precisely extracting meaningful features from positional information, 3d-OT achieves high-resolution analysis of spatial omics data. Utilizing the framework of soft communication optimal transport, it effectively tackles challenges such as non-rigid alignment and inconsistent resolution during the alignment process. Furthermore, it introduces a novel evaluation metric, the Chamfer distance, to assess alignment quality.