单细胞数据分析流程和验证实验
2022-02-04 本文已影响0人
Hayley笔记
1. 分析流程
Basic algorithms
- Expression quantification and quality control
- Normalization (CPM/TPM + logarithm)
- Batch effect correction for integrating multiple datasets
- Feature selection (only keeping highly variable genes)
- Principle Component Analysis (acceleration + denoise)
- Non-linear dimensional reduction for visualization (tSNE/UMAP)
- Unsupervised clustering (K-means/community detection)
- Differential expression (one-vs-rest)
- Annotate clusters based on marker genes
Additional algorithms
- Supervised annotation
- Trajectory inference for continuous cell states
- RNA velocity analysis
- Cell-cell interaction analysis
- Deconvolution analysis for bulk-data
- Integrated analysis with other techniques (scATAC-seq, CITE-seq, spatial-seq, TCR/BCR-seq)