umis

Unique Molecular Identifiers (UMIs) are short, random nucleotide sequences added to individual nucleic acid molecules during library preparation. They tag original molecules before amplification, enabling distinction between true molecule counts and duplicates produced by PCR, which improves quantitative accuracy in sequencing experiments.

UMIs are typically incorporated into sequencing adapters or primers as degenerate bases, with a length of about 6 to 12 nucleotides. The aim is to provide high combinatorial complexity so that two distinct original molecules are unlikely to share the same UMI. After sequencing, reads are grouped by their alignment position and their UMI; counting the number of unique UMIs per locus yields an estimate of the original molecule count rather than the total read count.

Data processing for UMIs involves error correction to address sequencing or PCR-induced errors in UMI sequences. Methods may collapse similar UMIs within a defined distance or apply directional adjacency approaches. Software tools such as UMI-tools implement UMI-aware deduplication and error correction to improve accuracy in molecule counting.

Applications of UMIs span both RNA and DNA sequencing. In bulk and single-cell RNA-seq, UMIs help quantify transcript abundance while reducing PCR amplification bias. They are also used in targeted DNA sequencing and other assays where precise molecule counting is important, including detection of low-frequency variants and accurate copy-number estimation.

Limitations and design considerations include the potential for UMI collisions when library complexity is high or UMIs are too short, and the impact of sequencing or PCR errors on counts if not corrected. Typical guidance favors 8–12 nt UMIs with balanced base composition and appropriate downstream processing aligned to sequencing depth and experimental goals.