MSClust: A Multi-Seeds based Clustering algorithm for microbiome profiling using 16S rRNA sequence.

Area of Science:

• Microbiology
• Bioinformatics
• Computational Biology

Background:

• Next-generation sequencing has generated vast amounts of 16S rRNA data for microbiome profiling.
• Clustering short sequences into operational taxonomic units (OTUs) is crucial for data analysis.
• Existing OTU inference methods often struggle to balance accuracy with computational efficiency for large datasets.

Purpose of the Study:

• To develop a novel heuristic clustering method for accurate and efficient OTU inference.
• To address the trade-off between inference accuracy and computational cost in microbiome data analysis.

Main Methods:

• Proposed MSClust, a multi-seeds based heuristic clustering method for OTU inference.
• MSClust adaptively selects multiple seeds per cluster and employs a greedy clustering strategy.
• The method was evaluated against existing heuristic clustering approaches.

Main Results:

• MSClust demonstrated reduced memory usage compared to existing methods.
• The new method achieved better biological accuracy in OTU inference.
• MSClust maintained efficiency and scalability for large-scale microbiome data analysis.

Conclusions:

• MSClust offers a promising solution for accurate and efficient OTU inference.
• The method balances computational performance with biological relevance in microbiome profiling.
• MSClust is a valuable tool for researchers working with large 16S rRNA sequence datasets.