单细胞RNA数据整合哪家强？

Snipaste_2025-08-05_14-41-16.png

Article: 10.1038/s41592-021-01336-8, Nature Methods, 2022

Methods

  1. benchmarked 16 popular tools on 13 datasets.

  2. methods were evaluated with/without feature selection & data scaling.

  3. 14 metrics and cell-cycle variations were used for the integration performances.

benchmarking
datasets used

Results

  1. scANVI, Scanorama and scVI perform best for scRNA-seq

  2. scATAC-seq integration performance depends on feature space

  3. cell annotations are available -> scGen and scANVI

  4. scATAC-seq data -> Harmony and LIGER

  5. Smart-seq2 + 10X -> Scanorama

summary
©著作权归作者所有,转载或内容合作请联系作者
平台声明：文章内容（如有图片或视频亦包括在内）由作者上传并发布，文章内容仅代表作者本人观点，简书系信息发布平台，仅提供信息存储服务。

推荐阅读更多精彩内容