iSMNN: Batch Effect Correction for Single-cell RNA-seq data via Iterative Supervised Mutual Nearest Neighbor Refinement

iSMNN is a flexible, accurate and robust method for batch effect correction of scRNA-seq data. iSMNN performs batch effect correction via iterative supervised mutual nearest neighbor (MNN) refinement, that is, performing multiple rounds of MNN refining and batch effect correction. Specifically, in the current implementation, iSMNN encompasses two major steps: one optional cluster harmonizing step and the other batch effect correction step. With cell cluster label information, iSMNN iteratively searches MNNs within each harmonized cell type, and performs batch effect correction using the iSMNN function. With the help of iterative MNN refining, iSMNN has been demonstrated advantages in removing batch effect yet maximally retaining cell type specific biological features, in particular to the scenarios where different batches largely differ. For details, please refer to our tutorial.

Comments and suggestions are welcome, please e-mail Yuchen Yang at yyuchen@email.unc.edu or Yun Li at yunli@med.unc.edu.