# dbGVOSCC: a comprehensive database of genetic variation for systems genetics research on oral squamous cell carcinoma

## TL;DR

dbGVOSCC is a new database that compiles genetic variations related to oral squamous cell carcinoma, supporting research into its genetic causes and potential treatments.

## Contribution

dbGVOSCC provides a comprehensive, curated database of genetic variations in oral squamous cell carcinoma with integrated bioinformatics tools.

## Key findings

- The database includes 1,788 genetic variation entries from 400 studies and 106,079 clinical samples.
- PPI analysis identified 15 key genes like TP53 and EGFR involved in OSCC pathways.
- User testing showed high usability scores (SUS 88.75 and NPS 90).

## Abstract

Oral squamous cell carcinoma (OSCC) is a highly aggressive malignancy of the oral epithelium, marked by a high rate of lymph node metastasis and a profound negative impact on patients’ quality of life. Despite its severity, no routine screening program currently exists for OSCC. To address the genetic heterogeneity underlying OSCC, we have developed a database of genetic variation in oral squamous cell carcinoma (dbGVOSCC; http://www.sysbio.org.cn/dbGVOSCC/).

OSCC literature (1991–2024) was queried from PubMed and screened manually and via PubTator, following predefined inclusion/exclusion criteria. Entities and relations were extracted from qualifying articles and organized into tables. The database adopted a browser/server architecture using HTML and XAMPP. Front-end was built with HTML and CSS for web display; server-side used Apache for infrastructure, MySQL for data management, and PHP/JavaScript for backend-frontend integration. Bioinformatics included mapping genes to STRING (confidence >0.9), hub gene identification via PPI degree centrality, and GO/KEGG enrichment with clusterProfiler (FDR-corrected). Usability was assessed using SUS and NPS surveys.

dbGVOSCC comprises 1,788 somatic genetic variation entries from 400 original studies and 106,079 clinical samples, covering epimutations/methylations (329), SNPs (411), point mutations excluding SNP (258), indels (98), CNVs (348), LOH (28), one locus mutation, plus 333 unspecified mutations. We curated 817 biomarker-linked variations (diagnostic n=71, therapeutic n=175, prognostic n=291; 277 multi-application). PPI analysis highlighted 15 key genes (e.g., TP53, CTNNB1, AKT1, EGFR, PIK3CA). Enrichment implicated proliferation, adhesion/migration, p53/DNA damage response, and PI3K–Akt signaling. User testing showed SUS 88.75 (grade A) and NPS 90.

dbGVOSCC represents a robust and reliable knowledge base, offering clinicians and researchers an open-source platform for personalized genotype-phenotype association studies and systems genetics research into the mechanisms of OSCC.