Department of Oncology, Lausanne Branch, Ludwig Institute for Cancer Research, CHUV and University of Lausanne, 1011, Lausanne, Switzerland.
AGORA Cancer Research Center, 1005, Lausanne, Switzerland.
Nat Commun. 2024 Jan 29;15(1):872. doi: 10.1038/s41467-024-45240-z.
Batch effects in single-cell RNA-seq data pose a significant challenge for comparative analyses across samples, individuals, and conditions. Although batch effect correction methods are routinely applied, data integration often leads to overcorrection and can result in the loss of biological variability. In this work we present STACAS, a batch correction method for scRNA-seq that leverages prior knowledge on cell types to preserve biological variability upon integration. Through an open-source benchmark, we show that semi-supervised STACAS outperforms state-of-the-art unsupervised methods, as well as supervised methods such as scANVI and scGen. STACAS scales well to large datasets and is robust to incomplete and imprecise input cell type labels, which are commonly encountered in real-life integration tasks. We argue that the incorporation of prior cell type information should be a common practice in single-cell data integration, and we provide a flexible framework for semi-supervised batch effect correction.
单细胞 RNA-seq 数据中的批次效应对跨样本、个体和条件的比较分析构成了重大挑战。尽管经常应用批次效应校正方法，但数据集成通常会导致过度校正，并可能导致生物变异性的丧失。在这项工作中，我们提出了 STACAS，这是一种用于 scRNA-seq 的批量校正方法，它利用了对细胞类型的先验知识，在集成时保留生物变异性。通过一个开源基准，我们表明半监督 STACAS 优于最先进的无监督方法，以及 scANVI 和 scGen 等监督方法。STACAS 可很好地扩展到大型数据集，并且对常见于实际集成任务中的不完整和不精确的输入细胞类型标签具有鲁棒性。我们认为，在单细胞数据分析集成中，应将先验细胞类型信息的纳入作为一种常见做法，并且我们提供了一个用于半监督批量效应校正的灵活框架。
