Logistic Regression Model Based On Liquid Biopsy Targeted Sequencing Technology For Cervical Cancer

Purpose:In statistics,the logistic model is a widely used statistical model that uses logistic functions to simulate binary dependent variables or more complex extensions,often used in data mining,automatic disease diagnosis,and economic forecasting.For example,explore the risk factors that caus e disease,and predict the probability of disease occurrence based on risk factors.Tumor patients' cf DNA has DNA fragments derived from tumors and metastatic lesions.Targeted sequencing of cf DNA in tumor patients can achieve low cost,easy access,non-invasive,dynamic disease detection,and now used to detect gene mutations,copy number and structural variation carried by patients,to further guide the patient's medication,radiotherapy and chemotherapy.While cf DNA is a DNA fragment that is free in plasma.These DN A fragments are not random and of equal length.The distribution of histones and the binding of transcription factors affect the free pattern of cf DN A.Here,we try to use the logistic model to treat patients and health.The DN A fragment free patterns of human WBC and cf DNA were compared and analyzed to study its biological significance and clinical application value.Method:A total of 153 samples of WBC and cf DN A in 57 patients with cervical cancer,3 tissue samples from 3 cervical cance r patients and WBC and cf DNA samples from 2 other healthy individuals,using 2158 amplicon panel involving 48 gene exons.The panel performs targeted sequencing based on multiplex PCR technology to obtain the sequencing depth of each amplicon.The original depth is first filtered and a series of normalized to obtain comparable amplicon depth.Then we look at the consistency of the depth of the amplicon between different samples,by the following methods: to see whether the depth ranking order of amplicons is consistent in all samples,visualize the amplicon depth using PCA,and systematically cluster the samples.We then tested 6 healthy human samples,including 6 WBC and 6 cf DNA samples,and also used 2158 amplicon panel for targeted sequencing.They were compared with 10 WBC and 17 cf DNA samples(including multiple samples)from the same batch of 10 cervical cancer patients.The standardized data of the samples were compared using a PCA and lasso logistic model for comparison of WBC and cf DNA samples.Finally,some genes involved in the model are explained biologically: the expression of these genes in healthy human,the comparison of genes in healthy human and tumor patient models,which metabolic pathways are involved in these genes in healthy people,and whether these genes are associated with tumor development in tumor patients.Result:The free pattern of cf DNA fragments in healthy individuals is related to the activity of transmembrane receptor protein tyrosine kinases,phosphorylation of metabolic processes,active regulation of stimuli,and positive regulation of cellular communication,which is consistent with whole blood in healthy human databases;the free pattern of cf DN A fragments in cervical cancer patients is also related to the activity of transmembrane receptor protein tyrosine kinase,protein phosphorylation,and in addition to the regulation of viral transcription of tumors,gene regulation in tumor hypoxia,positive regulation of intracellular signal transduction,regulation of growth,regulation of cell proliferation.Conclusion:The free fragments of cf DNA in cervical cancer patients and healthy people are inconsistent.This inconsistency reflects tumor epigenetic and genomic variation and changes and regulation of related metabolic and signaling pathways.This may be a direction for future cancer research.It can provide a new idea for the diagnosis and treatment of cervical cancer and even the diagnosis and treatment of other tumors.