Figure 2From: Unsupervised binning of environmental genomic fragments based on an error robust selection of l-mersThe pipeline of our binning algorithm First, the l-mer frequencies of each input sequence are counted. Then based on a novel modified Chebychev distance, a subset of l-mers is selected to create a feature vector. After that, k-mean clustering algorithm is applied to classify the fragments into taxonomic specific groups.Back to article page