A scalable assembly-free variable selection algorithm for biomarker discovery from metagenomes

BMC Bioinformatics

Table 1 Coverage biases in metagenome assemblies

Abundance class (Bin #)	Bin abundance level	% Binned reads in assembly	Estimated number of genomes in bin
Bin I	320	2.8 %	> = 3
Bin II	180	2.9 %	> = 1
Bin III	90	12.6 %	> = 2
Bin IV	30	54.0 %	> = 6
Bin V	9	0.5 %	> = 3

Unassembled (raw) reads derived from a xenobiotic degrading bacterial consortium (Chaussonerie et al. 2016 under review) were segregated by the AB-Cl module (k-mer size = 25) into 5 abundance classes (bins). Mapping of reads from individual bins on the metagenome assembly based on all the raw reads reveals a significant under-representation of abundance classes I, II, III and V

ISSN: 1471-2105