Accurate estimation of molecular counts from amplicon sequence data with unique molecular identifiers

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Amplicon sequencing is vital for genetic population studies, but accuracy is compromised by PCR and sequencing errors.
• Unique Molecular Identifiers (UMIs) help correct abundance estimates, yet existing methods struggle with UMI reuse, collisions, and error correction.

Purpose of the Study:

• To introduce DAUMI, a probabilistic framework for accurate detection and abundance estimation of true biological amplicon sequences.
• To address limitations in current UMI-based methods by accounting for UMI collisions and errors.

Main Methods:

• Developed DAUMI, a probabilistic framework utilizing unique molecular identifiers (UMIs).
• Implemented algorithms to recognize UMI collisions, even for similar sequences.
• Integrated methods for detecting and correcting PCR and sequencing errors within UMIs and sequences.

Main Results:

• DAUMI accurately detects true biological sequences and estimates their deduplicated abundance.
• The framework successfully handles UMI collisions and corrects errors in UMIs and sequences.
• DAUMI demonstrates superior performance compared to existing UMI-aware clustering methods on simulated and real data.

Conclusions:

• DAUMI provides a robust solution for accurate amplicon sequence abundance estimation.
• The framework enhances the reliability of genetic variant analysis by mitigating amplification and sequencing biases.
• DAUMI offers improved accuracy for exploring genetic heterogeneity and rare variants in populations.