Fig. 5

Candidate ICD-associated hub genes that can differentiate low and high parasitaemia were identified by protein‒protein interaction (PPI) network construction and machine learning. A The top 10 hub genes amongst genes upregulated in the ICD subtype 1 group were identified based on the maximal clique centrality (MCC) method using the cytoHubba plugin in Cytoscape. The color spectrum from red to yellow indicates the genes' connection degree, where darker hues signify a higher degree and greater gene importance. B The top 10 hub genes amongst genes downregulated in the ICD subtype 1 group were identified based on the MCC method using the cytoHubba plugin. C Binomial deviance was revealed by the LASSO regression model in the tenfold cross validation. The vertical dotted lines indicate the optimal values identified using the minimum and 1-SE criteria. D LASSO coefficient profiles of 20 selected ICD-associated genes in the tenfold cross-validation. E Twenty ICD-associated genes were ranked based on the importance of each gene associated with parasitaemia calculated by using RF machine learning algorithms. F Venn diagram showing the two candidate hub genes identified via both of the above algorithms