Fig. 1

CLAME methodology. Stage 1) read alignment: the metagenome is composed by reads from different genomes (represented by the red and green colors); each read, represented by a single rectangle, is aligned against all the reads; an adjacency list shows all the alignments for each read. Stage 2) edges analysis: The graph representation indicates the relation of the reads; the reads that belong to a shared region can connect the subgroups (the green reads are connected to the green reads by the relation between read 1 and read 2); these connections usually make the number-of-edges histogram depart from a normal like form; then the histogram helps the user to set the number-of-edges thresholds on a range, in which a normal distribution is observed; It allows users to make bins with reads belonging to a normal-like connection profile. Stage 3) graph traversal and bin generation: the bins are generated by traveling the graph and reporting each subgraph (e.g. {1, 9, 6, 5, 8} green reads and {0, 3, 7, 4} red reads)