UMI-nea: a fast, robust tool for reference-free UMI deduplication and accurate quantification.

Area of Science:

• Genomics and Bioinformatics
• Molecular Biology Techniques

Background:

• Unique Molecular Identifiers (UMIs) are crucial for correcting PCR bias and duplicates in high-throughput sequencing (DNA-seq, RNA-seq).
• Accurate UMI deduplication is essential for reliable quantification but challenging with error-prone long reads or ultra-high-depth short reads.
• Existing tools often use suboptimal clustering or Hamming distance, leading to inaccurate UMI groupings.

Purpose of the Study:

• To develop an improved UMI deduplication tool for accurate UMI grouping.
• To enhance quantification accuracy in high-throughput sequencing applications.
• To provide an efficient and robust solution for diverse sequencing data.

Main Methods:

• Introduced UMI-nea, a tool employing Levenshtein distance for UMI comparison.
• Implemented a novel clustering approach optimized for multithreading.
• Incorporated a data-guided adaptive UMI filter.

Main Results:

• UMI-nea achieved more accurate UMI groupings compared to three other indel-aware tools.
• Demonstrated efficient run time and robust performance across various sequencing platforms, depths, and UMI lengths.
• Enhanced quantification accuracy through its advanced filtering mechanism.

Conclusions:

• UMI-nea offers a superior method for UMI deduplication, improving quantification accuracy.
• The tool is effective across a wide range of sequencing conditions.
• UMI-nea provides a valuable advancement for high-throughput sequencing data analysis.