Fig. 2

Machine learning screening in NETs signature. (A) Heatmap shows the TOP 20 differentially expressed genes in NETs signature and their representative ssGSEA scores. (B) SVM-RFE screening of best variable genes in NETs signature. (C) Random forest screening for variable genes with weights greater than 0.7 in NETS signature. (D) Vene plot shows the NETs genes of Unicox analysis, Random Forest Analysis, and SVM-RFE analysis. (E) HRs for NETs genes. (F) Box plot demonstrating the difference in expression of NETs genes between CRC and normal tissue