目的 基于临床研究和单细胞测序技术分析脓毒症早期血小板数量和功能的变化。 方法 前瞻性病例对照研究，收集并比较不同预后组脓毒症患者疾病早期临床数据的差异；采用多因素logistic回归分析死亡的独立危险因素并通过ROC曲线评估临床指标的预测效能；招募健康志愿者和脓毒症患者采集外周静脉血样，分选细胞样本进行单细胞测序（sc-RNA seq）；借助生物信息学技术分析脓毒症早期血小板数量变化、显著差异的基因及其功能分析。 结果 （1）共224例患者入组，90 d存活率70.5%；与存活组相比，死亡患者疾病早期血小板计数和平均动脉压显著降低，血乳酸值和SOFA评分显著升高。（2）基于单细胞测序技术外周血细胞可注释分为6群，脓毒症早期固有免疫细胞（中性粒细胞，单核细胞和树突状细胞）数量占比较健康组显著升高（2.15∶1）；血小板则显著下降（0.31∶1）。（3）生物信息学技术鉴定出CD41/CD42a/CD61为血小板特异性分子，脓毒症时表达量显著升高，联合注释血小板能够独立成群。（4）脓毒症时血小板内771种基因表达显著上调，1 101种基因表达显著下调，包括参与细胞黏附、趋化作用和免疫应答等病理机制的核心分子；功能分析提示差异基因富集于凝血级联、免疫应答和细胞死亡，参与免疫细胞趋化、铁死亡以及焦亡相关的NOD样受体信号通路。 结论 脓毒症早期患者外周血小板数量减少且与不良预后相关；疾病早期血小板内CD41/CD42a/CD61显著升高可作为特异性标志分子，血小板不仅介导细胞黏附和凝血级联，还参与免疫细胞趋化、炎性应答和细胞病理性死亡等功能变化。

Objective We systematically analyze the changes in the count and function of platelet at the early sepsis based on clinical study and single cell sequencing. Methods Clinical data of sepsis patients at the early stage were collected and had been compared between different prognostic groups in the prospective case?control study. The independent risk factors of death were analyzed by logistic regression， and the predictive efficacy of clinical indicators was evaluated by receiver operating characteristic （ROC） curve. The healthy volunteers and sepsis patients were recruited. Clinical researchers collected peripheral venous blood samples for sorting cell samples to carry out single?cell RNA sequencing （sc?RNA seq）. Through bioinformatics techniques， we analyzed the changes in platelet count， the significantly differential?expressed genes and its enriched functional signaling pathways in the early stages of sepsis. Results （1） A total of 224 patients were enrolled， with a 90 day survival rate of 70.5%. Compared with the survival group， the count of platelet and MAP in the death group at the early stage of sepsis were significantly lower， but the plasma lactate content and SOFA score were significantly higher. （2） Based on single cell sequencing technology， cells are annotated as six groups. The proportion of innate immune cells （neutrophils， monocytes， and dendritic cells） was significantly increased in the early stage of sepsis compared to the healthy volunteers （2.15∶1）， while platelets significantly decreased （0.31∶1）. （3） Through bioinformatics technology， CD41/CD42a/CD61 was identified as platelet specific molecules， with significantly increased expression levels in sepsis. Three molecules can distinguish platelets together. （4） 771 genes were significantly upregulated and 1101 genes were significantly downregulated in platelets of patients with sepsis， including core molecules involved in physiological functions such as cell adhesion， chemotaxis， and immune response. Functional analysis suggests that differentially expressed genes are enriched in coagulation， immune functions and cell death， participating in oxidative phosphorylation， leukocyte chemotaxis， iron death， and NOD like receptor signaling pathways. Conclusion Reduced platelet count is associated with poor prognosis in the early stage of sepsis. The specific high expression molecules CD41/CD42a/CD61 that are significantly upregulated in platelets can serve as biomarkers for platelets. Platelets not only mediate cell adhesion and coagulation cascade， but also participate in functional changes such as immune cell chemotaxis， inflammatory response， and the pathological death of inflammatory cells.