Improving metagenomic binning results with overlapped bins using assembly graphs.
Vijini G Mallawaarachchi1, Anuradha S Wickramarachchi1, Yu Lin2
1School of Computing, College of Engineering and Computer Science, Australian National University, Canberra, Australia.
Algorithms for Molecular Biology : AMB
|May 5, 2021
Summary
GraphBin2 refines metagenomic binning by using assembly graph information to assign contigs to multiple species. This novel approach improves accuracy and handles shared genomic sequences, outperforming previous methods.
Area of Science:
- Microbial genomics
- Bioinformatics
- Computational biology
Background:
- Metagenomic sequencing enables microbial community analysis without culturing.
- Contigs from sequencing are binned into species-specific clusters.
- Existing tools assign contigs to only one bin, limiting accuracy for shared sequences.
Purpose of the Study:
- To introduce GraphBin2, a tool for refining metagenomic contig binning.
- To enable assignment of contigs to multiple species bins.
- To improve the accuracy of metagenomic binning.
Main Methods:
- Utilizes connectivity and coverage information from assembly graphs.
- Refines existing binning results.
- Infers contigs shared across multiple species.
Main Results:
- GraphBin2 successfully refines contig binning results.
- The tool assigns contigs to multiple bins, accommodating shared sequences.
- Experimental results show improved performance over existing tools on simulated and real data.
Conclusions:
- GraphBin2 enhances metagenomic binning by integrating coverage information into assembly graphs.
- The tool accurately detects and assigns contigs belonging to multiple species.
- GraphBin2 demonstrates superior performance compared to its predecessor, GraphBin.
