UMIche: a UMI-centric analysis platform for enhancing molecular quantification accuracy in bulk and single-cell sequencing

Abstract Unique molecular identifiers (UMIs) are widely utilised in RNA-sequencing (RNA-seq) technologies to eliminate polymerase chain reaction (PCR) duplicates, thereby improving the accuracy of molecular quantification. However, synthesis errors in UMIs during sequencing often compromise the reliability of UMI counting. While various computational methods have been developed for error correction, the lack of a unified benchmarking and validation framework impedes systematic evaluation and optimisation. Here, we present UMIche, an open-source, UMI-centric computational platform designed to enhance UMI-assisted molecular quantification. Its application to collapsing monomer and homotrimer UMIs across multiple bulk and single-cell sequencing protocols demonstrates effective PCR-duplicate removal. Across a broad spectrum of simulated and experimental datasets with ground truth, most UMI collapsing methods integrated into UMIche perform similarly under low-error conditions, whereas set-cover optimisation and majority voting show greater robustness as the UMI error profile becomes more pronounced. Together, these capabilities enable scalable UMI collapsing, systematic benchmarking, and method development to improve molecular quantification.