Table 1 Summary of twelve original genome binner and three refinning genome binner
From: Evaluating metagenomics tools for genome binning with real metagenomic datasets and CAMI datasets
Genome binner | Parameters | Model | Version to validate | Publication | Last update | Resources |
|---|---|---|---|---|---|---|
MaxBin | k-mer frequencies, coverage, single-copy genes | Expectation-maximization, bin number estimated from single-copy marker gene analysis | 2.2.6 | 2014 | 2019 | |
MetaBat | 4-mer frequencies, coverage | Modified K-medoids algorithm | 1&2.13 | 2015 | 2020 | |
Groopm | coverage, contig’s length, tetranucleotide frequency | Two way clustering, Hough partitioning, self-organizing map | 2 | 2014 | 2017 | |
CONCOCT | k-mer frequencies, coverage | Gaussian mixture models, bin number determined by variable Bayesian | 1.0.0 | 2014 | 2019 | |
MyCC | k-mer frequencies, coverage (optional), universal single-copy genes | Affinity propagation | 1 | 2016 | 2017 | |
MetaWatt | tetranucleotide frequency, coverage | Firstly clustering by empirical relationship of the average standard deviation at tetranucleotide frequency mean, then employing interpolated Markov models | 3.5.3 | 2012 | 2016 | |
BMC3C | frequency variation of oligonucleotides, coverage, codon usage | Ensemble k-means, construct a weigh graph and partition it by Normalized cuts [49, 50] | \ | 2018 | 2018 | |
Binsanity | coverage, tetranucleotide frequency, percent GC content | Affinity propagation | 0.2.8 | 2017 | 2020 | |
Autometa | sequence homology, single-copy genes, 5-mer frequency, coverage, single-copy genes | Lowest common ancestor analysis, DBSCAN algorithm, supervised decision tree classifier recruite unclustered contigs | \ | 2019 | 2020 | |
COCACOLA | k-mer frequency, coverage, co-alignment, paired-end read linkage | K-means based on L1 distance, non-negative matrix factorization with sparse regularization, hierarchical clustering | \ | 2017 | 2017 | |
SolidBin-naive | single-copy mark genes, tetranucleotide frequencies, coverage, pairwise constraints | Semi-supervised spectral Normalized cut | 1.1 | 2019 | 2020 | |
Vamb | ​tetranucleotide frequencies, coverage | Variational autoencoders, iterative medoid clustering algorithm | 2.0.1 | 2018 | 2020 | |
DAS Tool | original binner output bin sets | Refine bins according shared contigs between two original binner results | 1.1.1 | 2018 | 2019 | |
MetaWrap | original binner output bin sets | Separating every pair of contigs in different bins, selecting the best bin sets according completion and contamination | 1.2.2 | 2018 | 2019 | |
Binning_refiner | original binner output bin sets, single-copy genes | Scoring bins based on single-copy genes and picking up high-score bins iteratively | 1.4.0 | 2017 | 2019 |