Metalign: efficient alignment-based metagenomic profiling via containment min hash.

Area of Science:

• Microbiology
• Bioinformatics
• Computational Biology

Background:

• Metagenomic profiling is essential for understanding microbial communities.
• Alignment-based methods offer accuracy but are computationally intensive.
• Existing tools struggle with balancing speed and precision in microbiome analysis.

Purpose of the Study:

• To introduce Metalign, a novel computational method for efficient and accurate alignment-based metagenomic profiling.
• To address the computational infeasibility of traditional alignment approaches.
• To improve the prediction of microbial presence and abundance in complex samples.

Main Methods:

• Developed Metalign, incorporating a novel containment min hash for reference database pre-filtering.
• Utilized both uniquely aligned and multi-aligned reads for abundance estimation.
• Evaluated performance on real and simulated metagenomic datasets.

Main Results:

• Metalign demonstrated high performance and accuracy in metagenomic profiling.
• The method achieved competitive running times across various datasets.
• Metalign was the sole method to maintain high performance and efficiency consistently.

Conclusions:

• Metalign provides an efficient and accurate solution for alignment-based metagenomic profiling.
• The novel approach overcomes computational limitations of traditional methods.
• Metalign is a valuable tool for microbiome analysis, offering reliable microbial abundance predictions.