卵巢癌是一种异质性疾病，具有不同的分子表型。我们采用细胞分化轨迹分析对卵巢癌进行了分子分型，并提出了预后风险评分模型。利用inferCNV提供的拷贝数变异信息，我们确认了恶性肿瘤细胞。然后，根据分化相关基因（DRGs）将卵巢癌样本分为四个亚型。在亚型之间存在着在生存率、临床特征、肿瘤微环境评分以及ICGs表达水平上的显著差异。根据九个DRGs，我们生成了一种预后风险评分模型（1年的AUC为0.749，3年的AUC为0.651）。然后，我们获得了一种包括风险评分和临床病理特征的预后变量组合的表格，并预测了1年、3年和5年的总体生存率。最后，我们利用建立的风险模型探究了免疫逃逸的一些问题。我们的研究说明了细胞分化对于预测卵巢癌患者预后的显著影响，并为卵巢癌治疗和潜在免疫治疗策略提供了新的见解。版权所有©2023 Chen, Ni, Mu, Ying, Zhang和Wang

Ovarian cancer is a heterogeneous disease with different molecular phenotypes. We performed molecular typing of ovarian cancer using cell differentiation trajectory analysis and proposed a prognostic risk scoring model. Using the copy number variation provided by inferCNV, we identified malignant tumor cells. Then, ovarian cancer samples were divided into four subtypes based on differentiation-related genes (DRGs). There were significant differences in survival rates, clinical features, tumor microenvironment scores, and the expression levels of ICGs among the subtypes. Based on nine DRGs, a prognostic risk score model was generated (AUC at 1 year: 0.749; 3 years: 0.651). Then we obtained a nomogram of the prognostic variable combination, including risk scores and clinicopathological characteristics, and predicted the 1-, 3- and 5-year overall survival. Finally, we explored some issues of immune escape using the established risk model. Our study demonstrates the significant influence of cell differentiation on predicting prognosis in OV patients and provides new insights for OV treatment and potential immunotherapeutic strategies.Copyright © 2023 Chen, Ni, Mu, Ying, Zhang and Wang.