Method of determining tumor marker based on transcriptome data

Abstract

The invention discloses a method of determining a tumor marker based on transcriptome data, comprising: (1) acquiring transcriptome data, including first and second transcriptome data which comprise mRNA, LncRNA and miRNA expression data of first and second individual sample respectively, wherein differences of the first and second individual samples include one of paired relative phenotypic features; (2) establishing regularization logic regression models wherein each individual has phenotypic feature relationship with expression of three RNAs, and using the models to regress expression data of the three RNAs respectively to obtain molecular regression coefficients of the three RNAs; (3) performing grid searching, and determining three RAN thresholds according to the molecular regression coefficients of the three RNAs; (4) comparing the molecular regression coefficient of each RNA to a corresponding threshold, and determining three RNA candidate markers; (5) mixing the three RNA candidate markers to obtain RAN mixture data, and replacing transcriptome data with the RAN mixture data to carry out the steps (2) to (4) so as to determine a tumor marker.

Description

technical field [0001] The present invention relates to the field of bioinformatics, specifically, the present invention relates to a method for determining tumor markers based on transcriptome data and a group of tumor markers. Background technique [0002] Every cell has a complex gene expression regulation system, which cooperates with each other to perform normal biological functions. For the study of complex disease biological systems, it is necessary to integrate experimental and computational methods to analyze multi-level regulatory relationship data, thereby discovering pathogenic mechanisms and promoting disease diagnosis and treatment. Studies have found that the abnormal expression of some genes in tumor tissue is closely related to the occurrence and development of tumors, and thus become important tumor markers. In addition, some non-coding RNAs (such as microRNA, lncRNA, etc.) also have important regulatory functions in life activities, mediating and particip...

AI Technical Summary

Problems solved by technology

[0005] The existing method of predicting the risk of tumor prognosis based on a single type of molecular marker has certain limitations, because tumors have strong heterogeneity, and the internal regulatory system is very complex, and different types of molecules operate in different It is difficult to find molecular markers that can accurately predict disease prognosis based on a single type of omics data

A single type of molecular marker can only partially reflect some features within the tumor

Images