Welcome to scAtlas’s documentation!

Introduction

ScAtlas is a pipeline for building large-scale single-cell atlases. While there is exponential growth in the amount of single-cell data generated, a high quality reference atlas can be valuable resource for research groups to derive insights from their data. Here, we hope to provide a pipeline for building large-scale single-cell atlases that can be used for a variety of applications.

Contents

Contents:

• Introduction to scAtlas
  • Why atlas
• Preprocess
  • Metadata Collection
  • First round preprocess
  • Droplet discovery
  • First round annotation
  • Normalization and post normalization
  • Batch division
• Pre-integration
  • Integration methods benchmarking
  • Hyperparameter tuning
• Integration
  • First-level annotation
• Annotation
  • Step 0 : Leiden Clustering in Mesenchyme
  • Step 1: Choose the optimal cluster levels
  • Step 2: MRTree
  • Step 3: Marker Detection
  • Step 4: Prune Tree
  • Step 5: Annotation
  • Step 6: Final Visualization of Cell Type Clusters
  • Details of Each Level Annotation
• Trajectory
  • Trajectory Reduction
  • Stemness and Start Point Inference
  • Connectivity inference with PAGA
  • Pseudotime Inference
  • Generalize pseudotime with LightGBM
  • Trajectory Analysis of Gene Expression using tradeSeq
  • Statistical Background of TrajConserve
• References