Font Size: a A A

Screening Of Asthma Biomarkers Based On Statistical Machine Learning Methods And Studies Of Related Immune Infiltration

Posted on:2022-12-06Degree:MasterType:Thesis
Country:ChinaCandidate:S X ChenFull Text:PDF
GTID:2494306758985909Subject:Automation Technology
Abstract/Summary:PDF Full Text Request
Asthma is a common non-communicable chronic respiratory disease.At present,there are at least 300 million asthma patients in the world,accounting for about 4.3%of the global population.Asthma is a serious threat to human health,causing about 400,000 deaths every year.However,there are problems such as too high cost and too much consumption in the treatment methods based on clinical experiments in the process of prognosis and treatment of asthma.Therefore,the application of statistical machine learning methods to find biomarkers of asthma and explore the molecular mechanism of asthma has become a research hotspot.In this paper,we firstly studied the human asthma gene expression profile data in the GEO database by differential gene screening method,and obtained 271 differentially expressed genes related to asthma infection status.Then,through enrichment analysis,it was found that the differentially expressed genes were mainly enriched in immune-related functions and pathways.Next,through the core gene screening of the differential genes of asthma,68 highly significantly related genes in the process of asthma expression were obtained.On this basis,LASSO and SVM-RFE methods were used to initially screen asthma biomarkers,and 13 candidate asthma biomarkers were obtained.Through the control study of the data set,we finally determined two biomarkers,ADORA3 and CD24.Finally,we estimated the relative proportion of immune cells in asthma samples,and found that the infiltration of 7 types of immune cells was significantly different between the asthma group and the healthy group,and found that there was a significant correlation between asthma biomarkers and some immune cells.Although many scholars have studied asthma biomarkers,there are still many problems to be solved.Based on statistical machine learning methods,this paper focuses on the screening of asthma biomarkers and the relationship between asthma and immune infiltration.On the one hand,this paper provides a prognostic direction for the target screening and identification of asthma,on the other hand,discusses the molecular mechanism of asthma and its relationship with immune cells,and further provides a new scientific basis for the treatment of asthma.
Keywords/Search Tags:Asthma, immune-infiltration, biomarkers, LASSO, SVM-RFE
PDF Full Text Request
Related items