Integration for single-cell RNA sequencing data based on the shared cell type assignment

摘要

At present, many single-cell experiments from different laboratories and different sequencing platforms generate a large amount of RNA sequencing data, which is convenient for the research of new cell types detecting, cell clustering, gene regulatory network constructing and other downstream analysis. However, data from different laboratories or different sequencing platforms will contain batch effects that may compromise the integration and interpretation of the data. Existing methods don't achieve satisfactory integration results when the cell type composition varies greatly between batches. In this paper, we propose a novel method based on utilizing biological prior knowledge to guide the correction, which effectively solves the above problems and also outperforms other algorithms when similar cell types exist among batches or quantity distribution of cells from various cell types is seriously unbalanced.