Fig 1.
Overall working flowchart of promising potential biomarkers for NSCLC between USA cohort and Taiwan cohort.
GEO: Gene expression omnibus; DEGs: Differentially expressed genes; LIMMA: Lnear models for microarray data; PPI: Protein-protein interaction; ML: Machine learning; TCGA: The cancer genome atlas; 1DCNN: One-dimensional convolutional neural network; AUC: Area under the curve.
Table 1.
Description of the datasets for USA and Taiwan cohort.
Fig 2.
Volcano plot of the DEGs: (a) USA Cohort and (b) Taiwan cohort.
Table 2.
GO analysis for carcinoma associated DEGs for USA cohort.
Top 5 items were selected.
Table 3.
KEGG pathway analysis for carcinoma associated DEGs for USA cohort.
Top 5 items were selected.
Table 4.
Degree of connectivity for the selected core hub genes in the USA cohort.
Fig 3.
Analysis of PPI network and identification of core hub genes for USA cohort: (a) PPI network analysis for the carcinoma associated genes; (b) Identification of hub genes among four measures (Degree, Betweenness, Closeness, and MNC); and (c) PPI network analysis of identifying core hub genes.
Fig 4.
PPI network analysis of the hub genes for USA cohort.
Fig 5.
Identification of core genes for USA cohort by intersecting genes, obtained from PPI network, Module analysis, and ML-based approach.
The depth of the color corresponds to the quantity of hub genes present.
Fig 6.
Survival analysis of 12 core genes for USA cohort: (a) CLDN18; (b) AGER; (c) EDNRB; (d) TOP2A; (e) MME; (f) SPP1; (g) LPL; (h) SFTPD; (i) HMMR; (j) SFTPA1; (k) ANLN; (l) BIRC5.
The x-axis represents time to event (in days) and the y-axis represents survival probability.
Fig 7.
ROC curves and heatmap of six genes on test set for USA cohort (a) ROC curves and (b) Heatmap.
Fig 8.
Identification of most promising potential biomarkers for NSCLC by intersecting genes, obtained from Taiwan cohort and USA cohort.
Fig 9.
Regulatory network between potential biomarkers with TFs and miRNAs.
The red, magenta and green color nodes represent the potential biomarkers, TFs, and miRNAs, respectively. (a) Potential biomarkers and TF interaction network, (b) Potential biomarkers and miRNA interaction network.
Fig 10.
Potential biomarkers and drugs interaction netweork.
The red and green color nodes represent the potential biomarkers and drugs. (a) EDNRB and Drugs interaction and (b) MME and Drugs interaction.