RNA sequencing data from 521 COAD samples and associated clinical information were obtained from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). Another RNA sequencing data of 698 COAD samples and clinical information were also included for validation. RNA sequencing data were converted from fragments per kilobyte per million (FPKM) to the transcripts per million reads (TPM) format, and compared according to the corresponding clinicopathological information. As all data collected was publicly available, informed consent and ethical approval were not necessary to obtain.

To clarify the association between SNHG7 expression and clinical features of COAD, Wilcoxon signed-rank test and logistic regression were performed. The detailed clinicopathological characteristics of the patients with COAD are listed in Table 1 .

To assess the predictive potential of SNHG7 for COAD diagnosis, SNHG7 expression in COAD and normal tissues was compared using receiver operating characteristic (ROC) analysis. COAD and corresponding normal tissue data were obtained from the TCGA database. The analysis was performed using the R package “pROC” (version1.17.0.1), and the visualization was achieved using “ggplot2” (version 3.3.3).

Kaplan–Meier analysis, and univariate and multivariate Cox regression analyses were used for prognosis analysis. Nomograms were created using the R packages “rms” (version 6.2-0) and “survival” (version 3.2-10). R (v3.6.3; R Foundation for Statistical Computing, Vienna, Austria) was used to conduct all statistical studies, with p-values below 0.05 deemed significant.

COAD gene expression data in the HTSeq-TPM format were obtained from TCGA for analysis. SNHG7 coexpressed genes were screened using Pearson correlation coefficients (|r| > 0.4 and p < 0.001). To explore the possible biological functions and signaling pathways affected by SNHG7, the R package “cluster Profiler” was used to perform GO and KEGG analyses of coexpressed genes, with p < 0.05 deemed statistically significant. GO analysis included biological processes (BP), cell composition (CC), and molecular function (MF).

GSEA is a computational method used to determine whether an a priori defined gene set exhibits statistically significant and consistent differences between two biological states (16). In the present study, we elucidated the survival differences between groups with high and low SNHG7 expression using GSEA. Gene set permutations were performed 1,000 times for each analysis. The expression of SNHG7 was used as the phenotypic label. The nominal p-value and normalized enrichment score (NES) were used to identify the pathways enriched for each phenotype.

Based on TCGA data analysis, we found that SNHG7 was upregulated in 17 of the 33 cancer types investigated, including cholangiocarcinoma (CHOL), prostate adenocarcinoma (PRAD), and thyroid carcinoma (THCA) ( Figure 1A ). Further analysis showed that SNHG7 expression was much higher in patients with COAD than that in normal tissues (p < 0.001, Figure 1B ). These findings indicate that SNHG7 may play a significant regulatory role in the progression of COAD.

Clinical information, including sex, age, race, T stage, N stage, residual tumor, perineural invasion, lymphatic invasion, OS, and disease-specific survival (DSS) ( Table 1 ), for 521 COAD patients was obtained from TCGA database. SNHG7 expression was not only significantly correlated with race (p < 0.05) and residual tumor (p < 0.05), but was also closely correlated with OS (p < 0.05) and DSS (p < 0.01). No correlation was observed between SNHG7 expression and the other clinicopathological characteristics.

ROC curves were used to evaluate the potential of SNHG7 expression to identify patients with COAD. SNHG7 expression had high sensitivity and specificity for COAD diagnosis, with an area under the curve (AUC) of 0.912 (95% confidence interval [CI], 0.878–0.947) ( Figure 2A ). Further analysis showed that SNHG7 expression could diagnose T and N stages, with AUC values of 0.913 (95% CI: 0.878–0.947) and 0.921 (95% CI: 0.887–0.955), respectively ( Figures 2B, C ). Taken together, these results suggested that SNHG7 expression could represent a valuable tool to diagnosis COAD.

Kaplan–Meier analysis was used to confirm whether SNHG7 expression could be used for the prediction of clinical outcomes among patient with COAD. Overall, high SNHG7 expression was associated to shorter OS (hazard ratio [HR]: 1.85, p = 0.002) and DSS (HR: 2.35, p = 0.001) in COAD patients ( Figures 3A, B ). To determine whether SNHG7 expression had a predictive value for clinical outcomes, we also performed univariate and multivariate Cox regression analyses. As shown in Table 2 , SNHG7 expression (HR: 1.847, 95% CI: 1.244–2.741, p = 0.002) was an independent risk factor for OS. Age, T stage, pathological stage, lymphatic invasion, and carcinoembryonic antigen (CEA) level also showed prognostic advantages for clinical outcomes. In addition, the N stage showed prognostic advantages in both univariate (HR: 4.051, 95% CI: 2.593–6.329, p < 0.001) and multivariate (HR: 6.048, 95% CI: 1.006–36.361, p = 0.049) Cox regression analyses. Furthermore, SNHG7 expression, T stage, N stage, pathological stage, perineural invasion, lymphatic invasion, and CEA level were all independent risk factors for DSS ( Table 3 ). N stage (HR: 2.933, 95% CI: 0.218–39.407, p = 0.021) was an independent risk factor for DSS.

Based on the significant prognostic factors identified in the Cox regression analysis, prognostic nomograms were designed. Age, pathological stage, perineural or lymphatic invasion, and SNHG7 expression were included in the nomogram to predict OS (C-index = 0.836) ( Figure 3C ) and DSS (C-index = 0.875) ( Figure 3D ). These results indicated that SNHG7 expression was not only significantly upregulated in COAD but also had prognostic value, suggesting that SNHG7 has important regulatory functions in this type of cancer.

To screen for coexpressed genes of SNHG7, Pearson correlation coefficients were set as |r| > 0.4 and p < 0.001. The top 20 positively and negatively correlated coexpressed genes of SNHG7 are displayed in the form of a heatmap ( Figure 4 ).

Next, we performed GO and KEGG analysis of SNHG7- associated DEGs in COAD. GO analysis demonstrated that genes of GO-BP terms were significantly enriched in viral gene expression, viral transcription, establishment of protein localization to the endoplasmic reticulum (ER), and protein targeting to the ER. For GO-CC terms, the genes were mainly located in the cytosolic part, ribosome, ribosomal subunit, and the cytosolic ribosome. In GO-MF analysis, genes were enriched in the structural constituent of ribosomes, 5′–3′ RNA polymerase activity, DNA-directed 5′– 3′ RNA polymerase activity, and RNA polymerase II activity ( Table 4 ; Figure 5A ). As shown in Figure 5B ; Table 5 , KEGG pathway analysis indicated that the top pathways were mainly associated with Huntington’s disease, ribosome, amyotrophic lateral sclerosis, spliceosome, and RNA polymerases.

To further clarify biological functions of SNHG7 in COAD, GSEA enrichment analysis was performed on the high and low expression datasets of SNHG7. Significant differences (false discovery rate < 0.25, adjusted p < 0.05) were observed in the enrichment of the MSigDB Collection (c2.cp.v7.2.symbols.gmt). The most markedly enriched signaling pathways were screened based on their NES ( Figures 6A, B ). The results illustrated that lupus erythematosus (NES = −1.628, adjusted p = 0.048) and reactome cellular senescence (NES = −1.413, adjusted p = 0.048) were mainly enriched in the highly expressed SNHG7 phenotype.

To validate the prognostic robustness and clinical reproducibility of SNHG7, an independent cohort available at TCGA database, comprising 698 samples, was also analyzed. As shown in Figure 7A , SNHG7 expression was significantly upregulated in the tumor group as compared with normal tissues (p < 0.001). Similarly, ROC curve analysis also indicated that SNHG7 had a very high diagnostic value in COAD (AUC = 0.911, 95% CI: 0.879−0.943) ( Figure 7B ). Analysis of the OS ( Figure 7C ; HR: 1.55, p = 0.014) and DSS ( Figure 7D ; HR: 2.04, p = 0.002) of these patients further suggested that high SNHG7 expression was correlated with poor prognosis in COAD. These results were consistent with the conclusions of SNHG7 in cohort of 521 samples, indicating that the diagnostic and prognostic value of SNHG7 in COAD is credible and reproducible.