Identification of Novel Biomarkers for Predicting Prognosis and Immunotherapy Response in Head and Neck Squamous Cell Carcinoma Based on ceRNA Network and Immune Infiltration Analysis

Figure 6

Clinical correlation analysis. (a, b) ANLN and KIF23 expression was associated with tumor grade according to TISIDB datasets. (a) ANLN and (b) KIF23. (c–f) Key mRNAs among ceRNAs related to tumor stage according to TISIDB. (c) ITGA5, (d) KDELC1, (e) NFIA, and (f) PTX3. (g–i) Distribution of the expression of key ceRNAs across immune subtypes according to TISIDB: (g) ANLN, (h) KDELC1, (i) KIF23, and (j) NFIA. (k–q) Distribution of key ceRNA expression across molecular subtypes (TISIDB): (k) CFL2, (l) ITGA5, (m) KDELC1, (n) NFIA, (o) PTX3, (p) RELT, and (q) TMC7. Statistical significance of differential expression evaluated using Kruskal–Wallis test. C1 (wound healing), C2 (IFN-gamma dominant), C3 (inflammatory), C4 (lymphocyte depleted), and C6 (TGF-β dominant). Molecular subtypes include four types, namely, atypical (AT), basal (BA), classical (CL), and mesenchymal (MS) based on biological characteristics of genes highly expressed in each subtype.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

(p)

(q)