MetaBCC-LR: metagenomics binning by coverage and composition for long reads

Area of Science:

• Microbiology
• Bioinformatics
• Genomics

Background:

• Metagenomics studies reveal microbial community composition and structure.
• Binning is essential for characterizing microbial species in metagenomic data.
• Short-read sequencing limits current binning tools; long-read technologies offer solutions.

Purpose of the Study:

• To introduce MetaBCC-LR, a scalable, reference-free binning method for long reads.
• To overcome limitations of existing metagenomic binning tools for long reads.

Main Methods:

• MetaBCC-LR clusters long reads based on k-mer coverage histograms and oligonucleotide composition.
• It is a scalable, reference-free approach.

Main Results:

• MetaBCC-LR outperforms state-of-the-art reference-free binning tools on simulated and real long-read datasets.
• Achieved ~13% improvement in F1-score and ~30% improvement in ARI.
• Enhances long-read assembly quality while reducing computational costs.

Conclusions:

• MetaBCC-LR offers an efficient and accurate solution for long-read metagenomic analysis.
• Its performance paves the way for more effective applications in microbial community studies.