Probing as a new technique to assess single-cell RNA-seq batch correction

Single-cell RNA sequencing (scRNA-seq) is vulnerable to batch effects that could confound biological interpretation. We introduce the Batch Probing Score (BPS), a supervised metric quantifying residual batch signal after correction. Unlike existing unsupervised metrics, BPS directly measures the remaining batch signal that could affect downstream analysis. Evaluated on six datasets and six correction methods, BPS has the highest sensitivity among existing metrics used to evaluate batch correctors.