Skip to main content
Fig. 5 | Journal of Translational Medicine

Fig. 5

From: Rapid diagnosis and comprehensive bacteria profiling of sepsis based on cell-free DNA

Fig. 5

The performance of species inference based on the bacteria co-occurrence network. The curve shows the average recovery rate. For each testing sepsis sample, we performed 1000 trials. In each trial, we randomly removed 10–80% of observed bacterial species then inferred the presence of missing species from the co-occurrence network. The x-axis represents the removal percentage. a The y-axis represents the percentage of inferred species that were removed in the cross-validation. b The y-axis represents the total percentage of identified species for the cross-validation, including both inferred species and those that were never removed. c The y-axis represents the percentage of inferred species that were removed in for the validation based on an independent data. d The y-axis represents the total percentage of identified species for the validation of an independent data

Back to article page