基于单细胞多组学数据整合分析的帕金森病免疫机制研究

研究表明帕金森病（Parkinson’s Disease，PD）可能是一种自身免疫性疾病，深入研究其免疫机制对PD临床诊疗有重要意义。近年来单细胞测序技术在肿瘤、阿尔茨海默病等复杂疾病研究中取得巨大成功，然而尚未应用在PD研究中。本项目团队通过对PD患者和正常人进行单细胞测序，发现PD患者外周血中存在疾病特异的T细胞亚群。然而，现有单细胞组学大数据分析工具和软件方法不足以全面揭示PD免疫系统的复杂性。为此，本项目拟从PD基因组、单细胞转录组和免疫组数据出发，优化和开发有针对性的整合、降噪和聚类算法，构建精细的PD免疫图谱，整合贝叶斯推断、隐马尔可夫模型等筛选PD特异淋巴细胞亚群以及TCR/BCR序列，提出基于深度学习PD新生抗原预测模型和算法，多组学数据整合构建单细胞层面基因调控模型，并结合生物学实验进行验证。研究成果不仅为揭示PD免疫机制奠定基础，也将可能推动PD免疫疗法研发。

Studies have shown that Parkinson's disease (PD) may be an autoimmune disease, in-depth study of its immune mechanism is of great significance for the clinical diagnosis and treatment of PD. In recent years, single cell sequencing technology has achieved great success in the research of tumor, Alzheimer's disease and other complex diseases, but it has not been applied in PD research. By sequencing the single cells of PD patients and normal people, our team found that there were disease-specific T cell subsets in the peripheral blood of PD patients. However, the existing big data analysis tools and software methods are not enough to fully reveal the complexity of PD immune system. Therefore, this project plans to optimize and develop targeted integration, noise reduction and clustering algorithms based on PD’s genome, single cell transcriptome and immune group data, build a fine PD immune map, integrate Bayesian inference, hidden Markov model, etc. to screen PD specific lymphocyte subsets and TCR / BCR sequences, and propose a new PD antigen prediction model and algorithm based on deep learning. Gene regulation model on single cell level was constructed based on multi-omics data integration and validated by biological experiments. The research results will not only lay a foundation for revealing the immune mechanism of PD, but also promote the development of immunotherapy of PD.