Abstract We describe the integration of multi-batch cytometry datasets (iMUBAC), a flexible, robust, and scalable computational framework for unsupervised cell-type identification across multiple batches of high-dimensional cytometry datasets. After overlaying cells from healthy controls across multiple batches, iMUBAC learns batch-specific cell-type classification boundaries and identifies aberrant immunophenotypes in patient samples. We illustrate unbiased and streamlined immunophenotyping, using both in-house and public mass and flow cytometry datasets.

Multi-batch cytometry data integration for optimal immunophenotyping