Fig. 1

The cgMSI workflow for strain identification on nanopore metagenomic data using two-stage MAP estimation. cgMSI starts by mapping the core gene alleles of the target species to raw reads (using reads as reference) and selects candidate strains using MAP probability estimation. After that, cgMSI maps the aligned reads to the full reference genomes of the candidate strains and identifies the target strain using the second-stage MAP probability estimation. The Monte Carlo method is used to estimate the proportion of the target strain reads containing a complete allele of a core gene, which is further used to estimate the coverage of the target strain