Establishment and bioinformatics analysis of an early prediction model for parkinson′s disease based on gut microbiota
-
摘要:
目的 探究基于肠道菌群对人群未来患帕金森症(parkinson′s disease, PD)的早期预测模型进行构建与评价，并对肠道菌群宏基因KO组进行功能分析，探讨PD的潜在治疗靶点。 方法 基于Zenodo数据库，一方面对肠道菌群相对丰度数据进行标准分数标准化及ZicoSeq降维，采用基于自适应最小绝对收缩和选择算法(least absolute shrinkage and selection operator, LASSO)变量选择的logistic回归算法建立预测模型，使用受试者工作特征(receiver operating characteristic, ROC)曲线下面积(area under curve, AUC)和校准曲线评价模型预测效能，采用临床决策曲线(decision curve analysis, DCA)进行临床使用价值的评价；另一方面，对肠道菌群宏基因KO组数据使用limma包鉴定差异表达基因(differentially expressed genes, DEGs)，对DEGs进行基因本体(gene ontology, GO)和京都基因与基因组百科全书(Kyoto encyclopedia of genes and genomes, KEGG)分析。通过结合蛋白质相互作用网络(protein-protein interaction networks, PPI)、支持向量机-递归特征消除(support vector machine-recursive feature elimination，SVM-RFE)和随机森林(random forest, RF)对DEGs进行筛选。 结果 基于自适应LASSO变量选择的logistic回归分析模型的ROC曲线以及校准曲线显示模型预测效果良好。DCA结果显示模型净收益较大。通过PPI网络分析及机器学习方法，最终筛选出6个核心DEGs，即阿拉伯糖转运系统渗透蛋白(L-arabinose operon Q, araQ)、甘油醛-3-磷酸脱氢酶，Ⅱ型、dCTP脱氨酶(dCTP deaminase, dcd)、颗粒19 kDa蛋白(signal recognition particle 19, SRP19)、加工前体5，核糖核酸酶P/MRP亚基(芽殖酵母)[processing of precursor 5, ribonuclease P/MRP subunit (S. cerevisiae), POP5]、肌醇-3-磷酸合酶1(inositol-3-phosphate synthase 1, ISYNA1)。 结论 基于自适应LASSO变量选择的logistic回归分析模型对PD的预测具有优势，从而实现对PD患者的早发现、早干预、早治疗；相关核心基因的发现为PD的治疗提供科学指导和帮助。 Abstract:Objective To explore the construction and evaluation of an early prediction model for parkinson′s disease (PD) in the population based on gut microbiota to conduct functional analysis of gut microbiota macro-genus KO groups to explore potential therapeutic targets for PD. Methods Gut microbiota relative abundance data from the Zenodo database were standardized using Z-Score and dimensionality reduction was performed using ZicoSeq. An adaptive least absolute shrinkage and selection operator (LASSO) binary logistic regression algorithm was employed to establish the prediction model. The performance of the model was evaluated using the area under the receiver operating characteristic (ROC) curve and calibration curve, and clinical utility was assessed using decision curve analysis (DCA). Differential expression genes (DEGs) in gut microbiota macro-genus KO groups were identified using the limma package. Gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) analyses were performed on DEGs. DEGs were further screened using protein-protein interaction networks (PPI), support vector machine recursive feature elimination (SVM-RFE), and random forest (RF). Results The ROC curve and calibration curve of the adaptive LASSO binary logistic regression model showed good predictive performance. The DCA curve showed a significant net benefit of the model. PPI network analysis and machine learning methods identified 6 core DEGs, namely L-arabinose operon Q (araQ), mitochondrial FAD-dependent glyceraldehyde-3-phosphate dehydrogenase, dcd, SRP19, POP5, and ISYNA1. Conclusions The adaptive LASSO binary logistic regression algorithm model has significant advantages in predicting PD, enabling early detection, intervention, and treatment of PD patients. The discovery of relevant core genes provides scientific guidance and assistance for the development of PD treatments. -
Key words:
- Parkinson′s disease /
- Gut microbiota /
- Machine learning /
- Bioinformatic analysis
-
图 1 基于自适应LASSO变量选择的logistic回归的变量选择
A：基于自适应LASSO变量选择的logistic回归分析模型的log(λ)与自变量数目对应曲线；上横坐标为不同log(λ)对应的模型中非零系数自变量的个数；B：基于自适应LASSO变量选择的logistic回归分析模型中每条曲线代表每个自变量系数的变化轨迹；上横坐标为不同log(λ)对应的模型中非零系数自变量的个数; a: 平均均方误差最小时最优调和系数; b: 平均均方误差一个标准误内最优调和系数。
Figure 1. Variable selection based on adaptive LASSO binary logistic
A: the log (λ) curve corresponding to the number of independent variables in the adaptive LASSO binary logistic regression model; The upper horizontal axis represents the number of non-zero coefficient independent variables in the model corresponding to different logs (λ); B: the change trajectory of each independent variable coefficient represented by each curve in the adaptive LASSO binary logistic regression model; The upper horizontal axis represents the number of non-zero coefficient independent variables in the model corresponding to different logs (λ); a: the optimal harmonic coefficient with the minimum mean square error; b: the mean square error is the best harmonic coefficient within one standard error.
图 2 帕金森症早期预测模型的评价指标
A：帕金森症早期预测模型受试者工作特征曲线图；B：帕金森症早期预测模型校准曲线图；C：帕金森症早期预测模型决策分析曲线图。
Figure 2. Evaluation indicators for early parkinson′s disease prediction models
A: the receiver operating characteristic curve of the parkinson′s disease early prediction model; B: the calibration curve of the early parkinson′s disease prediction model; C: the decision analysis curve of the early parkinson′s disease prediction model.
图 4 通过机器学习筛选核心基因及其差异表达
A: SVM-RFE的十折交叉验证曲线；B: RF的袋外错误率与特征数量关系图；C: 挑选出的所有核心DEGs的差异表达箱线图; a表示组与组之间的显著性差异P值＜0.001，b表示组与组之间的显著性差异P值＜0.01。
Figure 4. Screening core genes through machine learning and their differential expression
A: the ten fold cross validation curve of SVM-RFE; B: the relationship between out of bag error rate and feature quantity in RF; C: the differential expression box plots of all selected DEGs; a represents a significant difference between groups, with a P < 0.001, b represents a significant difference between groups with a P < 0.01.
表 1 与帕金森症高度相关的9个协变量
Table 1. covariates highly related to parkinson′disease
变量
Variable帕金森症组
Parkinson′s group ①(n=435)正常组
Normal group ① (n=219)t/χ2值 value 帕金森症组与正常组对比
Parkinson′s group versus normal groupOR值 value
(95% CI)P值 value 年龄/岁 Age/years 68.5±8.5 65.8±8.8 3.749 1.1(1.3~4.1) ＜0.001 男性 Male 274(63) 65(30) 64.729 4.0(2.8~5.8) ＜0.001 饮酒 Drinking 160(37) 112(51) 12.365 0.6(0.4~0.8) ＜0.001 泻药 Laxative 131(30) 23(11) 31.124 3.7(2.2~6.2) ＜0.001 益生菌 Probiotics 53(12) 41(19) 5.059 0.6(0.4~0.9) 0.033 疼痛 Pain 100(23) 35(16) 4.366 1.6(1.0~2.5) 0.041 抑郁 Depression 165(38) 50(23) 15.051 2.1(1.4~3.1) ＜0.001 抗组胺药 Antihistamine 80(16) 70(32) 21.810 0.4(0.3~0.6) ＜0.001 安眠药 Sleep aid 175(40) 58(26) 12.001 1.9(1.3~2.7) ＜0.001 注：①以人数(占比/%)或x±s表示。
Note：①Number of people(proportion/%) or x±s.
下载: 导出CSV
表 2 DEGs在GO分析和KEGG分析中主要富集的前10个通路
Table 2. The top ten pathways that DEGs is mainly enriched in GO analysis and KEGG analysis
GO分析 GO analysis 通路 Pathway 生物学过程 Biological process 细胞质翻译 Cytoplasmic translation 基底切除修复、间隙填充 Based-excision repair, gap-filling 翻译终止 Translation termination 翻译启动 Translation initiation 翻译延伸 Translation elongation 以DNA为模板复制 DNA-templated DNA replication 细胞核DNA复制 Nuclear DNA replication 碱基切除修复 Base-excision repair 细胞周期DNA复制 Cell cycle DNA replication DNA复制 DNA replication 细胞成分 Cellular components 胞质核糖体 Cytosolic ribosome 胞质核糖体大亚基 Cytosolic large ribosomal subunit 核糖体亚基 Ribosomal subunit 核糖体大亚基 Large ribosomal subunit 核糖体 Ribosome 黏着斑 Focal adhesion 细胞-基底结 Cell-substrate junction 复制叉 Replication fork 多聚核糖体 Polysome 线粒体基质 Mitochondrial matrix 分子功能 Molecular functions 翻译因子活性，RNA结合 Translation factor activity, RNA binding 翻译调节活性，核酸结合 Translation regulator activity nucleic acid binding 核糖体的结构成分 Structural constituent of ribosome 翻译延伸因子 Translation elongation factor activity 核糖体结合 Ribosome binding 核糖核蛋白复合物结合 Ribonucleoprotein complex binding 分子功能 Molecular functions 翻译调节活性Translation regulator activity 翻译起始因子活性 Translation initiation factor activity 催化活性，作用于 DNA Catalytic activity, acting on DNA 镁离子结合 Magnesium ion binding KEGG富集分析 KEGG enrichment analysis DNA复制 DNA replication 核糖体 Ribosome 新冠肺炎病毒 Cornavirus disease-COVID-19 碱基切除修复 Base excision repair 错配修复 Mismatch repair 核苷酸切除修复 Nucleotide excison repair 丁酸代谢 Butanoate metabolism 缬氨酸、亮氨酸和异亮氨酸降解 Valine, lecucine and isoleucine degradation 非同源末端连接 Non-homologous end-joining 信使核糖核苷酸检测途径 mRNA surveilance pathway
下载: 导出CSV
-
[1] Valdes AM, Walter J, Segal E, et al. Role of the gut microbiota in nutrition and health[J]. BMJ, 2018, 361: k2179. DOI: 10.1136/bmj.k2179. [2] Nair AT, Ramachandran V, Joghee NM, et al. Gut microbiota dysfunction as reliable non-invasive early diagnostic biomarkers in the pathophysiology of parkinson's disease: a critical review[J]. J Neurogastroenterol Motil, 2018, 24(1): 30-42. DOI: 10.5056/jnm17105. [3] Sun MF, Shen YQ. Dysbiosis of gut microbiota and microbial metabolites in parkinson's disease[J]. Ageing Res Rev, 2018, 45: 53-61. DOI: 10.1016/j.arr.2018.04.004. [4] 凌茜文, 于洁琼, 彭思远, 等. 肠道微生物基因组学与代谢组学分析联合应用的研究进展[J]. 微生物学通报, 2021, 48(4): 1292-1301. DOI: 10.13344/j.microbiol.china.200585.Ling XW, Yu JQ, Peng SY, et al. Combined application of intestinal microbial genomics and metabolomics analysis: a review[J]. Microbiology China, 2021, 48(4): 1292-1301. DOI: 10.13344/j.microbiol.china.200585. [5] Lin H, Peddada SD. Analysis of microbial compositions: a review of normalization and differential abundance analysis[J]. NPJ Biofilms Microbiomes, 2020, 6(1): 60. DOI: 10.1038/s41522-020-00160-w. [6] Yang L, Chen J. A comprehensive evaluation of microbial differential abundance analysis methods: current status and potential solutions[J]. Microbiome, 2022, 10(1): 130. DOI: 10.1186/s40168-022-01320-0. [7] 田青. 改进的自适应Lasso算法及其在人口出生率中的应用[D]. 西安: 西安财经大学, 2023.Tian Q. Improved adaptive Lasso algorithm and its application to birth rate[D]. Xi'an: Xi'an University of Finance and Economics, 2023. [8] Lin XH, Li C, Zhang YH, et al. Selecting feature subsets based on SVM-RFE and the overlapping ratio with applications in bioinformatics[J]. Molecules, 2017, 23(1): 52. DOI: 10.3390/molecules23010052. [9] Diaz-Uriarte R, de Andres S A. Variable selection from random forests: application to gene expression data[EB/OL]. (2005-03-16)[2024-02-23]. DOI: https://arxiv.org/abs/q-bio/0503025v2 .[10] Poewe W, Seppi K, Tanner CM, et al. Parkinson disease[J]. Nat Rev Dis Primers, 2017, 3: 17013. DOI: 10.1038/nrdp.2017.13. [11] Dahodwala N, Shah K, He Y, et al. Sex disparities in access to caregiving in parkinson disease[J]. Neurology, 2018, 90(1): e48-e54. DOI: 10.1212/WNL.0000000000004764. [12] 常滢滢, 梅命珠, 于慧卿, 等. 心元胶囊对慢性心力衰竭心血瘀阻证患者心脏超声参数、脂质代谢及血液流变学的影响[J]. 现代生物医学进展, 2023, 23(5): 954-958. DOI: 10.13241/j.cnki.pmb.2023.05.031.Chang YY, Mei MZ, Yu HQ, et al. Effects of Xinyuan capsule on cardiac ultrasound parameters, lipid metabolism and blood rheology in patients with chronic heart failure with cardiovascular stasis[J]. Advances in Modern Biomedicine, 2023, 23(5): 954-958. DOI: 10.13241/j.cnki.pmb.2023.05.031. -
点击查看大图
计量
- 文章访问数: 33
- HTML全文浏览量: 38
- PDF下载量: 10
- 被引次数: 0