Head and neck squamous cell carcinoma (HNSCC) has been widely reported and considered as one of the most threatening diseases to human health. Derived from complicated tissue subtypes, HNSCC has diverse symptoms and pathogenesis. They make the identification of the core carcinogenic factors of such diseases at the multi-cell level difficult. With the development of single-cell sequencing technologies, the effects of non-malignant cells on traditional bulk sequencing data can be eliminated directly. On the basis of fresh single-cell RNA-seq data, we set up a computational filtering strategy for tumor cell identification in an expression rule manner. This strategy can reveal the accurate expression distinction between tumor cells and adjacent tumor microenvironment, which are all supported by literature reports. Validated by several independent datasets, these rule genes can further group HNSCC patients with significant difference on survival risks. Thus, the establishment of our computational approach may not only provide an efficient tool to identify malignant cells in the tumor ecosystem but also deepen our understanding of tumor heterogeneity and tumorigenesis.

中文翻译：

---

基于单细胞RNA测序的肿瘤特异性表达规则揭示了头颈部鳞状细胞癌的基本基因表达模式。

头颈部鳞状细胞癌（HNSCC）已被广泛报道，被认为是对人类健康威胁最大的疾病之一。HNSCC来自复杂的组织亚型，具有多种症状和发病机制。它们使得难以在多细胞水平上鉴定此类疾病的核心致癌因素。随着单细胞测序技术的发展，可以直接消除非恶性细胞对传统批量测序数据的影响。基于新鲜的单细胞RNA-seq数据，我们以表达规则的方式建立了用于肿瘤细胞鉴定的计算过滤策略。该策略可以揭示肿瘤细胞与相邻肿瘤微环境之间的准确表达区别，所有这些都得到文献报道的支持。这些规则基因已通过多个独立的数据集验证，可以进一步将HNSCC患者分组，其生存风险存在显着差异。因此，我们计算方法的建立不仅可以提供识别肿瘤生态系统中恶性细胞的有效工具，而且可以加深我们对肿瘤异质性和肿瘤发生的理解。