Skip to main content
Fig. 6 | BMC Bioinformatics

Fig. 6

From: SMILE: systems metabolomics using interpretable learning and evolution

Fig. 6

Interpretable learning results of SMILE on the AD metabolomic data. All sub-figures are generated by SMILE web interface. A Model accuracy graph shows the testing accuracy of all predictive models that use a chosen feature by clicking a bar in Fig. 5B. This graph, for instance, shows the testing accuracy of models that use cis-5,8,11,14-Eicosatetraenoic acid as an effective feature. B Detailed model figure provides symbolic representation of a selected model, by clicking a point in A. C Feature pairwise co-occurrence graph represents the frequencies using a heat map, and provides the metabolite names though mouse hover. D Two-feature scatter plot depicts the distribution of the two chosen metabolites’ concentrations. The metabolite pair shown in this figure has the highest co-occurrence frequency in the AD metabolomic data

Back to article page