Skip to main content
Fig. 2 | Journal of Translational Medicine

Fig. 2

From: Incorporating knowledge of disease-defining hub genes and regulatory network into a machine learning-based model for predicting treatment response in lupus nephritis after the first renal flare

Fig. 2

Characterization of transcriptomic profiles in LN. A PCA of genes between LN samples and healthy controls in GSE32591 and GSE112943. GSE32591 is divided according to tissue origins to evaluate the expression patterns in two tissue compartments. Aquamarine: LN samples. Red: healthy controls. B MA plots visualize the DEGs between LN and control samples. The data are transformed into M (logFC) and A (mean expression). Red dots indicate DEGs with |logFC ≥ 1|. Aquamarine dots indicate DEGs with |logFC < 1|. C Heatmaps of up- and down-regulated genes in LN. Clusters of genes are stratified using hierarchical clustering. Expression levels are z-transformed. Color bar indicates the transformed expression value. D Scatter plot of correlation between logFC of common DEGs in glomerular and tubulointerstitial compartments. Significance of correlation is performed by Spearman’s test. E Venn diagram of DEGs among two tissue compartments in GSE32591 and renal tissue in GSE112943. F Violin plots reveal differences in IFI44L expression between LN samples and controls in GSE32591 (left panel) and GSE112943 (right panel). Wilcoxon’s rank sum test P value is shown. G Dot plot reveals clustered enrichment ontology categories from ORA. –Log10-transformed multiple testing-adjusted P value is shown for each enriched term. H Dot plot reveals the significance of association with diseases via DisGeNET. Log10-transformed multiple testing-adjusted P value is shown for each enriched phenotype

Back to article page