Skip to main content
Figure 8 | BMC Bioinformatics

Figure 8

From: A systematic evaluation of high-dimensional, ensemble-based regression for exploring large model spaces in microbiome analyses

Figure 8

Comparison of inclusion probability profiles across all approaches. The top panels show the sorted inclusion probabilities plotted against the rank of the variables in the model. The bottom panels show the pairwise similarities between the inclusion probability profiles of all the approaches. The similarity between any pair of approaches i and j is given as \(S_{\textit {ij}}=\frac {1}{1+D_{\textit {ij}}}\), where D ij is the Euclidean distance between the inclusion probability vectors of approaches i and j. We generated the dendrogram using agglomerative clustering with the centroid linkage method. We then ordered the correlogram using the order of models generated in the dendrogram. a. Bacterial models, b. Fungal models.

Back to article page