Tumor-driven IncRNA screening method and system based on interaction and co-expression characteristics of three-dimensional genome

Abstract

The invention provides a tumor-driven IncRNA screening method based on interaction and co-expression characteristics of a three-dimensional genome. The method comprises the following steps: acquiring tumor information containing tumor TAD, expression quantity data, tumor-driven PCG and lncRNA gene information; calculating CDT and RNA-DNA interaction characteristics and co-expression characteristics according to the tumor information; according to the CDT and RNA-DNA interaction characteristics and the co-expression characteristics, identifying the tumor driving lncRNA, and according to the co-expression characteristics, classifying the tumor driving lncRNA; and the tumor driving lncRNA comprises a cancer promoting lncRNA, a cancer inhibiting lncRNA and a dual-effect lncRNA. According to the method, three-dimensional genome interaction and co-expression characteristics are utilized to realize tumor lncRNA recognition directly influencing tumor driven PCG expression, and the recognition efficiency, accuracy and precision are improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of biological information, in particular to a tumor driving IncRNA screening method and system based on three-dimensional genome interaction and co-expression characteristics. BACKGROUND

[0002] At present, potential tumor IncRNAs can be efficiently recognized by constructing a calculation method to analyze multi-omics data such as genomes and epigenomes. However, the existing methods are either based on only genomic data for analysis, with low sensitivity, or use expression or epigenetic characteristics in a single tissue to train a machine learning model for prediction, but these characteristics are not widely representative. Although there are some calculation methods that can identify tumor IncRNAs, they are only limited to tumor correlation, and lack effective differentiation and identification of driving IncRNAs. Therefore, it is very necessary to design a tumor driving IncRNA screening method and system based on three-dimensional genome interaction and co-expression characteristics. SUMMARY

[0003] The purpose of the application is to provide a tumor driving IncRNA screening method and system based on three-dimensional genome interaction and co-expression characteristics, which uses three-dimensional genome interaction and co-expression characteristics to realize the identification of tumor IncRNAs that directly affect the expression of tumor driving PCG, so as to improve the efficiency, accuracy and precision of identification.

[0004] To achieve the above purpose, the application provides the following scheme:

[0005] A tumor driving IncRNA screening method based on three-dimensional genome interaction and co-expression characteristics, comprising the following steps:

[0006] Obtain tumor information containing tumor TAD, expression data, tumor driving PCG and IncRNA gene information;

[0007] Calculate CDT, RNA-DNA interaction characteristics and co-expression characteristics according to the tumor information;

[0008] According to the CDT, RNA-DNA interaction characteristics and co-expression characteristics, tumor driving IncRNAs are identified, and the tumor driving IncRNAs are classified according to the co-expression characteristics; the tumor driving IncRNAs include: oncogenic IncRNAs, tumor suppressor IncRNAs and dual-acting IncRNAs.

[0009] Optionally, obtaining tumor information containing tumor TAD, expression data, tumor driving PCG and IncRNA gene information comprises:

[0010] Perform data analysis on the Hi-C sample of the target tumor type to obtain tumor TAD;

[0011] Obtaining expression data by RNA-seq analysis method for data analysis of tumor types;

[0012] Taking average operation on TUSON ranking in TUSON dataset, selecting top 150 tumor driving PCGs, and dividing the tumor driving PCGs into oncogenic PCGs, tumor suppressor PCGs and double-acting PCGs.

[0013] Optionally, the CDT, RNA-DNA interaction feature and co-expression feature are calculated according to the tumor information, including:

[0014] Taking union operation according to the interval completely containing the tumor driving PCG in the tumor TAD, obtaining the CDT;

[0015] Predicting whether the triplex site and R-loop site are formed between the lncRNA in the CDT and the tumor driving PCG, and if formed, integrating the triplex site and R-loop site into the RNA-DNA interaction feature;

[0016] Obtaining the co-expression feature by calculating the Spearman expression correlation coefficient between the lncRNA in the CDT and the tumor driving PCG.

[0017] Optionally, the calculation formula of the Spearman expression correlation coefficient is as follows:

[0018] ;

[0019] Wherein, is the expression value of the lncRNA in the CDT in the tumor sample, is the expression value of the tumor driving PCG in the tumor sample, and are the ranks of and respectively, and are the averages of the ranks.

[0020] Optionally, the tumor driving lncRNA is identified according to the CDT, RNA-DNA interaction feature and co-expression feature, and the tumor driving lncRNA is classified according to the co-expression feature, including:

[0021] If the expression of the tumor driving lncRNA is positively correlated with the oncogenic PCG or negatively correlated with the tumor suppressor PCG, it is determined as an oncogenic lncRNA;

[0022] ​​If the tumor driver lncRNA is negatively correlated with the expression of the oncogenic PCG or positively correlated with the expression of the tumor suppressor PCG, it is determined as a tumor suppressor lncRNA.

[0023] If the tumor driver lncRNA meets the classification conditions of the oncogenic lncRNA and the tumor suppressor lncRNA, it is determined as a dual-acting lncRNA.

[0024] Optionally, the positive correlation determination condition is that the Spearman expression correlation coefficient is greater than 0.5; and the negative correlation determination condition is that the Spearman expression correlation coefficient is less than -0.5.

[0025] A tumor driver IncRNA screening system based on three-dimensional genomic interaction and co-expression characteristics comprises:

[0026] A data acquisition module is configured to acquire tumor information comprising tumor TAD, expression data, tumor driver PCG, and lncRNA gene information.

[0027] A feature extraction module is configured to calculate CDT, RNA-DNA interaction characteristics, and co-expression characteristics according to the tumor information.

[0028] An lncRNA recognition module is configured to recognize tumor driver lncRNA according to the CDT, RNA-DNA interaction characteristics, and co-expression characteristics, and classify the tumor driver lncRNA according to the co-expression characteristics; the tumor driver lncRNA comprises an oncogenic lncRNA, a tumor suppressor lncRNA, and a dual-acting lncRNA.

[0029] According to the specific embodiments of the present application, the following technical effects are disclosed: the tumor driver IncRNA screening method based on three-dimensional genomic interaction and co-expression characteristics provided by the present application comprises: acquiring tumor information comprising tumor TAD, expression data, tumor driver PCG, and lncRNA gene information; calculating CDT, RNA-DNA interaction characteristics, and co-expression characteristics according to the tumor information; recognizing tumor driver lncRNA according to the CDT, RNA-DNA interaction characteristics, and co-expression characteristics, and classifying the tumor driver lncRNA according to the co-expression characteristics; the tumor driver lncRNA comprises an oncogenic lncRNA, a tumor suppressor lncRNA, and a dual-acting lncRNA. The method uses three-dimensional genomic interaction and co-expression characteristics to realize the recognition of tumor lncRNA that directly affects the expression of tumor driver PCG, and improves the efficiency, accuracy, and precision of the recognition. BRIEF DESCRIPTION OF DRAWINGS

[0030] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a flowchart of the tumor-driven incRNA screening method of the present invention. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0034] like Figure 1 As shown, this invention provides a tumor-driven incRNA screening method based on three-dimensional genome interaction and co-expression characteristics, comprising the following steps:

[0035] Step 100: Obtain tumor information including tumor TAD, expression level data, tumor driver PCG and lncRNA gene information;

[0036] Step 200: Calculate CDT, RNA-DNA interaction characteristics, and co-expression characteristics based on tumor information;

[0037] Step 300: Identify tumor-driving lncRNAs based on CDT, RNA-DNA interaction characteristics, and co-expression characteristics, and classify tumor-driving lncRNAs according to co-expression characteristics; tumor-driving lncRNAs include: oncogenic lncRNAs, tumor suppressor lncRNAs, and dual-function lncRNAs.

[0038] Specifically, in the present embodiment, in step 100, existing tumor topologically associated domain (TAD) data is collected for a specific tumor type; if there is no TAD data, a Hi-C sample of the specific tumor type is collected, and data analysis is performed using the HiCExplorer tool to obtain tumor TAD data. For a specific tumor type, existing expression data is collected; if there is no existing expression data, RNA-seq data of the specific tumor type is analyzed using the RNA-seq analysis method to obtain expression data. The specific analysis process is described at the website https: / / ngdc.cncb.ac.cn / lncexpdb / pipeline, and will not be described again in the present embodiment. The tumor driver PCG is obtained from the CGC dataset and the TUSON dataset. In some embodiments, the tumor driver PCG is divided into oncogenic PCG, tumor suppressor PCG and dual-acting PCG according to the existing identification in the CGC dataset; in other embodiments, for a single tumor type, the TUSON scores of PCG in pan-cancer and single tumor type are averaged, and the top 150 PCGs are taken as the oncogenic PCG, tumor suppressor PCG and dual-acting PCG in the single tumor type. All lncRNA gene information is obtained according to the existing reference genome. In the present embodiment, the lncRNA gene information is extracted from the gtf file provided by LncBook, and the lncRNA gene converted from GRCh38 version to GRCh37 version in v1.9 version is used, a total of 94810 lncRNA genes.

[0039] Specifically, the implementation process of step 200 in the present embodiment is as follows:

[0040] S201: According to the tumor TAD data and the tumor driver PCG information, the TAD completely containing the tumor driver PCG in each sample is obtained, and the TAD intervals completely containing the tumor driver PCG in these samples are taken to obtain the cancer driver TAD (CDT), and the lncRNA completely falling in the CDT is recorded.

[0041] S202: The embodiment uses Triplexator and QmRLFS-finder software to predict whether a lncRNA in the CDT and a tumor-driving PCG in the CDT form a triplex and an R-loop site, respectively, and only retains the lncRNA predicted to have an RNA-DNA (triplex or R-loop) binding site. Before prediction, the sequence of the lncRNA and the PCG is first obtained from the hg19.fa file using the getfasta function in BEDTools. Specifically, the prediction process of the triplex is as follows: input the sequence of the lncRNA and the PCG, and use the default parameters to predict the number of triplex sites formed between the lncRNA and the PCG; the prediction process of the R-loop is as follows: first, predict the R-loop structure formed by each lncRNA, and then match the number of complementary pairs in the PCG according to the predicted RIZ sequence, and count it as the number of R-loop sites formed between the lncRNA and the PCG.

[0042] S203: Obtain the co-expression characteristics by calculating the Spearman expression correlation coefficient between the lncRNA and the tumor-driving PCG in the CDT , and the calculation formula is:

[0043] ;

[0044] wherein, is the expression value of the lncRNA in the tumor sample , is the expression value of the tumor-driving PCG in the tumor sample , and are the ranks of and , respectively, and are the averages of the ranks. When the absolute value of the Spearman expression correlation coefficient is greater than 0.5, it is considered to be a co-expression relationship, and therefore the embodiment only retains the lncRNA whose absolute value of the Spearman expression correlation coefficient is greater than 0.5.

[0045] Specifically, step 300 obtains the tumor-driving lncRNA according to all the above characteristics. If the expression of the tumor-driving lncRNA is positively correlated with the expression of the oncogenic PCG or negatively correlated with the expression of the tumor-suppressive PCG, it is determined to be an oncogenic lncRNA; if the expression of the tumor-driving lncRNA is negatively correlated with the expression of the oncogenic PCG or positively correlated with the expression of the tumor-suppressive PCG, it is determined to be a tumor-suppressive lncRNA; if both the oncogenic and the tumor-suppressive cases are included, it is determined to be a dual-acting lncRNA.

[0046] Further, the positive correlation determination condition is Spearman expression correlation coefficient > 0.5; the negative correlation determination condition is Spearman expression correlation coefficient < -0.5.

[0047] The application also provides a tumor driving IncRNA screening system based on three-dimensional genomic interaction and co-expression characteristics, comprising:

[0048] A data acquisition module is configured to acquire tumor information comprising tumor TAD, expression data, tumor driving PCG and IncRNA gene information.

[0049] A feature extraction module is configured to calculate CDT, RNA-DNA interaction features and co-expression features according to the tumor information.

[0050] An IncRNA identification module is configured to identify tumor driving IncRNA according to CDT, RNA-DNA interaction features and co-expression features, and classify the tumor driving IncRNA according to the co-expression features; the tumor driving IncRNA comprises cancer-promoting IncRNA, cancer-inhibiting IncRNA and dual-acting IncRNA.

[0051] The application has the following beneficial effects:

[0052] 1) The method for mining tumor driving IncRNA can identify tumor IncRNA directly affecting tumor driving PCG expression by using three-dimensional genomic interaction and co-expression characteristics, thereby improving the efficiency, accuracy and precision of identification.

[0053] In the specification, each embodiment is described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and the same or similar parts between each embodiment can be mutually referred to.

[0054] In the application, specific examples are used to describe the principles and implementation modes of the application, and the above embodiment descriptions are only used to help understand the method and core idea of the application; meanwhile, for those skilled in the art, according to the idea of the application, the specific implementation mode and application range can be changed. In summary, the content of the specification should not be understood as a limitation of the application.

Claims

1. A method for screening tumor-driving IncRNAs based on three-dimensional genomic interaction and co-expression signatures, characterized in that, The method comprises the following steps: obtaining tumor information containing tumor TAD, expression data, tumor driving PCG and lncRNA gene information; calculating CDT, RNA-DNA interaction feature and co-expression feature according to the tumor information; identifying tumor driving lncRNA according to the CDT, the RNA-DNA interaction feature and the co-expression feature, and classifying the tumor driving lncRNA according to the co-expression feature; the tumor driving lncRNA comprises: oncogenic lncRNA, tumor suppressor lncRNA and dual-acting lncRNA.

2. The method of tumor-driving IncRNA screening based on three-dimensional genomic interaction and co-expression signatures according to claim 1, characterized in that, The method for obtaining tumor information containing tumor TAD, expression data, tumor driving PCG and lncRNA gene information comprises: performing data analysis on a Hi-C sample of a target tumor type to obtain the tumor TAD; performing data analysis on the tumor type by an RNA-seq analysis method to obtain the expression data; averaging the TUSON rankings in the TUSON dataset, selecting the top 150 tumor driving PCGs, and dividing the tumor driving PCGs into oncogenic PCGs, tumor suppressor PCGs and dual-acting PCGs.

3. The method of tumor-driving IncRNA screening based on three-dimensional genomic interaction and co-expression signatures according to claim 2, characterized in that, The method for calculating CDT, RNA-DNA interaction feature and co-expression feature according to the tumor information comprises: performing a set operation according to the interval completely containing the tumor driving PCG in the tumor TAD to obtain the CDT; predicting whether a triplex site and an R-loop site are formed between the lncRNA in the CDT and the tumor driving PCG, and if so, integrating the triplex site and the R-loop site into the RNA-DNA interaction feature; obtaining the co-expression feature by calculating the Spearman expression correlation coefficient between the lncRNA in the CDT and the tumor driving PCG.

4. The method of tumor-driving IncRNA screening based on three-dimensional genomic interaction and co-expression signatures according to claim 3, characterized in that, The Spearman expression correlation coefficient The formula for calculating the Spearman expression correlation coefficient is: ； wherein, is the expression value of the lncRNA in the tumor sample, is the expression value of the tumor driver PCG in the tumor sample, and are the ranks of and respectively, and are the average of the ranks respectively.​​ 5. The method of tumor-driving IncRNA screening based on three-dimensional genomic interaction and co-expression signatures according to claim 4, characterized in that, The method for identifying tumor driving lncRNA according to the CDT, the RNA-DNA interaction feature and the co-expression feature, and classifying the tumor driving lncRNA according to the co-expression feature comprises: if the expression of the tumor driving lncRNA and the oncogenic PCG is positively correlated or the expression of the tumor driving lncRNA and the tumor suppressor PCG is negatively correlated, the tumor driving lncRNA is determined as the oncogenic lncRNA; if the expression of the tumor driving lncRNA and the oncogenic PCG is negatively correlated or the expression of the tumor driving lncRNA and the tumor suppressor PCG is positively correlated, the tumor driving lncRNA is determined as the tumor suppressor lncRNA; if the tumor driving lncRNA meets the classification conditions of the oncogenic lncRNA and the tumor suppressor lncRNA, the tumor driving lncRNA is determined as the dual-acting lncRNA.

6. The method of tumor-driving IncRNA screening based on three-dimensional genomic interaction and co-expression signatures according to claim 5, characterized in that, The positive correlation determination condition is that the Spearman expression correlation coefficient is greater than 0.5; and the negative correlation determination condition is that the Spearman expression correlation coefficient is less than -0.

5.

7. A tumor-driving IncRNA screening system based on three-dimensional genomic interaction and co-expression features, characterized in that, The method comprises: a data acquisition module configured to obtain tumor information containing tumor TAD, expression data, tumor driving PCG and lncRNA gene information; The feature extraction module is configured to calculate CDT, RNA-DNA interaction features and co-expression features according to the tumor information. The lncRNA identification module is configured to identify tumor-driving lncRNAs according to the CDT, the RNA-DNA interaction features and the co-expression features, and classify the tumor-driving lncRNAs according to the co-expression features; the tumor-driving lncRNAs include cancer-promoting lncRNAs, cancer-suppressing lncRNAs and double-acting lncRNAs.

Images