Brain glioma microenvironment formation key molecular mechanism analysis method

Abstract

The invention relates to the technical field of biological information, in particular to a brain glioma microenvironment formation key molecular mechanism analysis method, which comprises the following steps: calling expression data to analyze a cell marker sequence to construct a segmentation interval, calculating a candidate factor expression direction to judge trend consistency, identifying expression aggregation difference to construct a split node section, and constructing a subsection; according to the method, partitions are constructed on the basis of marker expression syn-position, judgment is carried out in combination with candidate factor expression directions and marker trends, factor collaboration features are constructed according to the number of trends consistent times, and the semantic sorting information is generated by analyzing the channel change trends to generate drift scores, evaluating multi-label output stability screening key factors and analyzing literature word order positions. Identifying and expressing an aggregation and split structure, guiding a functional pathway to perform trend analysis in a scoring interval and construct a dynamic trajectory, judging label stability and empowerment according to pathway scoring difference, and determining a factor semantic position in combination with a literature word order structure to realize integrated support of regulation and control information and literature evidence.

Description

TECHNICAL FIELD

[0001] The application relates to the field of bioinformatics, in particular to a method for analyzing key molecular mechanisms of brain glioma microenvironment formation. BACKGROUND

[0002] The field of bioinformatics includes a series of computing methods and system models for acquiring, processing, analyzing and interpreting biological big data. Based on molecular biology, high-throughput sequencing technology and information science, it focuses on revealing the information structure and interaction relationship at the molecular level in the biological body through computing means, and researches multiple aspects such as sequence alignment, gene expression profile analysis, gene regulation network modeling, protein structure prediction and omics data integration analysis. The purpose is to identify key nodes and functional modules in biological systems by using computing analysis tools. In practical applications, bioinformatics is often cross-fused with molecular biology, cell biology and mathematical modeling to realize the digital expression and modeling analysis of complex biological systems. The method for analyzing key molecular mechanisms of brain glioma microenvironment formation refers to using the expression matrix extracted from high-throughput sequencing data as the basis, combining biomarker databases for multiple molecular screening, identifying candidate molecular factors related to the microenvironment through co-expression analysis, further screening tumor-specific expression characteristics through differential analysis, and further identifying functional factors involved in microenvironment formation using co-enrichment scoring method and cross-verification with literature evidence database. The method is aimed at identifying regulatory factors involved in brain glioma microenvironment construction, and specifically relies on standardized transcriptome data, tumor typing labels, molecular interaction networks and functional annotation databases and other basic data sources to complete the systematic identification of key molecules through feature extraction, enrichment clustering and literature correlation comparison.

[0003] The traditional method for analyzing key molecular mechanisms of brain glioma microenvironment formation relies on static feature extraction of the expression matrix in the co-expression analysis and differential screening process, lacks dynamic description of the continuity of expression trend and local aggregation behavior, and thus cannot effectively detect path variation in cross-interval pathway regulation mode identification. In the correlation link based on literature verification, the potential indication value of semantic order to the regulation direction is not considered, resulting in that the factor role identification result lacks context structure support, and in the judgment of factors with large scoring discrepancy of multiple labels, there are problems such as lack of cross-label stability and fuzzy semantic association, which affect the screening reliability of key factors and the direction judgment accuracy of regulation relationship. SUMMARY

[0004] In order to solve the technical problems existing in the prior art, the method for analyzing key molecular mechanisms of brain glioma microenvironment formation provided by the embodiments of the application comprises the following steps: To achieve the above objectives, the present invention employs the following technical solution: a method for elucidating the key molecular mechanisms of glioma microenvironment formation, comprising the following steps: S1: Call the expression data of glioma tissue, analyze the expression order of cell markers in the sample sequence, construct segmented intervals, calculate the expression change direction of candidate factors in each interval, combine the expression trend of markers to make direction correspondence judgment, count the number of trend consistency, and construct behavioral synergy quantification results; S2: Using the behavioral collaborative quantification results, determine the expression clustering of candidate factors in continuous samples, combine expression differences and subtype distribution differences, identify expression splitting locations and construct branch structure markers, and generate trajectory splitting node structure segments; S3: Using the aforementioned trajectory splitting node structural segments, compare the expression levels of functional pathways within the immune score segment intervals, analyze pathway expression trends and interval changes, construct a dynamic trajectory index for pathway activation, and generate a functional pathway drift trend score. S4: Using the functional pathway drift trend score, compare the output differences of candidate factors under multiple labels, compare the contribution difference range between label outputs, determine the stability of the degree of difference and adjust the weight of the corresponding label, evaluate the degree of cross-label stability influence, identify key factors, and generate the core evaluation results of regulation. S5: Using the evaluation results of the aforementioned regulatory core, detect the word order structure of candidate factors and glioma terms in multiple articles, analyze the word order position distribution, establish semantic directional labels, and generate directional path semantic ranking information.

[0005] As a further aspect of the present invention, the behavioral coordination quantification results include the number of times the direction corresponds, the trend coverage range, and the expression change direction sequence; the trajectory splitting node structure segment specifically includes splitting region positioning information, subtype distribution comparison records, and node branch identifiers; the functional pathway drift trend score includes the pathway continuous change trend, the score interval span index, and enrichment direction change characteristics; the regulation core evaluation results specifically include cross-label contribution score, label stability coefficient, and key factor screening number; and the directional path semantic ranking information includes semantic preorder occurrence frequency, regulation direction distribution label, and path structure positioning index.

[0006] As a further aspect of the present invention, the constructed pathway activation dynamic trajectory index refers to recording the expression enhancement trend and the number of sustained intervals of the pathway in each interval based on the changes in the enrichment level of functional pathways in multiple immune score segments, reflecting the dynamic response characteristics of the pathway under various immune states, and measuring the degree of activity in microenvironment regulation.

[0007] As a further scheme of the present application, the establishing semantic directionality label refers to that based on the in-sentence arrangement order of the candidate factor and the brain glioma related term in the literature, the frequency proportion of being in the front or back position in the expression is counted, the semantic position in the regulation path is determined through the direction preference, and the upstream regulation and downstream response types are divided.

[0008] As a further scheme of the present application, the acquiring step of the behavior coordination quantization result is specifically: S101: acquiring brain glioma tissue expression data, detecting the expression arrangement order of the cell marker in the sample sequence, constructing expression segmented intervals in turn according to the expression value, calculating the expression mean value change direction of each marker in the corresponding segmented interval, and generating segmented expression change direction data; S102: based on the segmented expression change direction data, calculating the expression mean value change direction of the candidate factor in each segmented interval, comparing the expression change direction of the candidate factor and each cell marker in the same segmented interval, recording the change direction consistent condition in each interval, and obtaining the trend corresponding statistical value; S103: calling the trend corresponding statistical value, counting the number of times that the change trend of the candidate factor and the cell marker is consistent in each segmented interval, and combining the interval distribution law to establish a numerical scale reflecting the overall expression behavior coordination of the candidate factor, and acquiring the behavior coordination quantization result.

[0009] As a further scheme of the present application, the acquiring step of the trajectory splitting node structure section is specifically: S201: based on the behavior coordination quantization result, judging the expression dense section of the candidate factor in the continuous sample, detecting the aggregation mode of the expression value in the sample arrangement, and screening the differential expression amplitude of the expression dense interval to generate an expression dense interval differential coefficient; S202: according to the expression dense interval differential coefficient, constructing an interval boundary point recognition condition, judging the expression breakpoint position, and combining the subtype distribution to which the samples on both sides of the expression breakpoint belong, calculating the proportion of the subtype distribution on both sides, and generating a boundary point subtype distribution coefficient; S203: calling the boundary point subtype distribution coefficient, performing group difference test, screening the position of the expression interval where the splitting behavior occurs, combining the splitting interval positioning with the branch information, establishing a node structure value, and acquiring a trajectory splitting node structure section.

[0010] As a further scheme of the present application, the acquiring step of the functional pathway drift trend score is specifically: S301: according to the trajectory splitting node structure section, acquiring the enrichment expression level of the immune score in each functional pathway in multiple continuous segmented intervals, analyzing the enrichment level change of each pathway in the scoring interval, and obtaining pathway interval enrichment mean value data; S302: Call the pathway interval enrichment mean data, compare the enrichment direction change of each functional pathway in the score interval, screen the pathway sequence with consistent enrichment direction in the score interval, count the continuous change interval of the pathway, and obtain the pathway continuous change interval value; S303: Based on the pathway continuous change interval value, analyze the continuity of the enrichment trend of each functional pathway in the continuous score interval and the interval change amplitude, construct the pathway activation dynamic trajectory index, and generate the functional pathway drift trend score.

[0011] As a further scheme of the present application, the obtaining step of the regulatory core evaluation result is specifically: S401: According to the functional pathway drift trend score, obtain the output data of each candidate factor under the three labels of immune score, prognosis score and subtype classification, analyze the contribution difference value of each label output, count the contribution difference interval of each factor, and obtain the label contribution interval value; S402: Based on the label contribution interval value, compare the contribution difference value range between the label outputs, judge the stability of the difference degree and adjust the weight of the corresponding label, and obtain the label output processing result; S403: Call the label output processing result, summarize the output of each candidate factor after weighting in all label combinations, evaluate the cross-label stable influence degree, establish the factor stable influence score, identify and screen the key factor, and obtain the regulatory core evaluation result.

[0012] As a further scheme of the present application, the obtaining step of the direction path semantic sorting information is specifically: S501: According to the regulatory core evaluation result, detect the arrangement order of the candidate factor name and the brain glioma related term in the sentence in multiple related literature paragraphs, record the relative order before and after each group of words, establish the order statistical table of the factor in the literature, and obtain the factor word order statistical data; S502: Based on the factor word order statistical data, analyze the number of times of the candidate factor before and after the brain glioma related term in multiple literature paragraphs, and calculate the word order distribution index; S503: Call the word order distribution index, judge the position of the target factor in the regulation chain, including the starting regulator and the response participant, summarize the semantic label and the word order distribution index of each candidate factor, establish the semantic direction label, and obtain the direction path semantic sorting information.

[0013] Compared with the prior art, the present application has the advantages and positive effects that: In the present application, based on the cell marker expression order in the sample sequence, the partition is constructed, the corresponding judgment is carried out combining the candidate factor expression change direction and the marker trend, the synergistic behavior between factors is described by the number of consistent trends, the expression aggregation and structure split characteristics are identified, the trend change analysis of the functional pathway in the continuous scoring interval is guided by the split node structure, the pathway dynamic trajectory is established, the stability judgment and weight summary of the contribution difference of the multi-label output are carried out based on the subsequent pathway score, the semantic position of the factor in the regulation chain is combined with the literature sequence structure, the association integration of the regulation information and the literature evidence is realized, and the structural accuracy of the regulation factor identification and the systematicness of the semantic direction judgment are enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0015] Figure 1 The step flowchart of the present application is shown in the figure. Figure 2 The S1 refinement schematic diagram of the present application is shown in the figure. Figure 3 The S2 refinement schematic diagram of the present application is shown in the figure. Figure 4 The S3 refinement schematic diagram of the present application is shown in the figure. Figure 5 The S4 refinement schematic diagram of the present application is shown in the figure. Figure 6 The S5 refinement schematic diagram of the present application is shown in the figure. DETAILED DESCRIPTION

[0016] The technical solutions in the present application will be described below with reference to the drawings.

[0017] In the embodiments of the present application, the words such as "example", "for example" and the like are used to represent as an example, illustration or description. Any embodiment or design scheme described as "example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0018] In the embodiments of the present application, "image" and "picture" can be used interchangeably, and it should be pointed out that their meanings are consistent when their differences are not emphasized.

[0019] In the embodiments of the present application, sometimes the subscript such as W1 can be written in the form of non-subscript such as W1, and their meanings are consistent when their differences are not emphasized.

[0020] In order to make the technical problems, technical solutions and advantages to be solved by the present application more clear, the following will be described in detail in combination with the drawings and specific embodiments.

[0021] Please refer to Figure 1 The embodiments of the present application provide a brain glioma microenvironment formation key molecular mechanism analysis method, comprising the following steps: S1: calling brain glioma tissue expression data, analyzing the expression arrangement order of cell markers in sample sequences, constructing a segmented interval, calculating the expression change direction of candidate factors in each interval, combining the marker expression trend to judge the direction correspondence, counting the number of consistent trends, and constructing behavior coordination quantization results; S2: using the behavior coordination quantization results, judging the expression aggregation of the candidate factors in the continuous samples, combining the expression difference and the subtype distribution difference, identifying the expression split position and constructing the branch structure marker, and generating the trajectory split node structure segment; S3: using the trajectory split node structure segment, comparing the expression level of the functional pathway in the immune score segmented interval, analyzing the pathway expression trend and the interval change, constructing the pathway activation dynamic trajectory index, and generating the functional pathway drift trend score; S4: using the functional pathway drift trend score, comparing the output difference of the candidate factors under multiple labels, comparing the contribution difference range between label outputs, judging the stability of the difference degree and adjusting the weight of the corresponding label, evaluating the cross-label stable influence degree, identifying the key factor, and generating the regulation core evaluation result; S5: using the regulation core evaluation result, detecting the syntactic structure of the candidate factor and the brain glioma term in multiple literatures, analyzing the syntactic position distribution, establishing the semantic directionality label, and generating the directional path semantic ordering information.

[0022] The behavior coordination quantification result includes the number of corresponding directions, the trend coverage range, the expression change direction sequence, the trajectory split node structure section, and the specific split area positioning information, the subtype distribution comparison record, and the node branch identification. The functional pathway drift trend score includes the pathway continuous change trend, the score interval span index, and the enrichment direction change characteristics. The regulatory core evaluation result includes the cross-label contribution score, the label stability coefficient, and the key factor screening number. The direction path semantic sorting information includes the semantic pre-sequence appearance frequency, the regulatory direction distribution label, and the path structure positioning index.

[0023] The construction of the pathway activity dynamic trajectory index refers to recording the expression promotion trend and the number of continuous intervals of the pathway in each interval based on the enrichment level change of the functional pathway in multiple immune score segments, reflecting the response dynamic characteristics of the pathway under multiple immune states, and measuring the activity degree in the microenvironment regulation.

[0024] The establishment of the semantic directionality label refers to counting the frequency proportion of being in the front or back position in the expression based on the in-sentence arrangement order of the candidate factor and the brain glioma related term in the literature, determining the semantic position in the regulatory path through the direction preference, and dividing into the upstream regulation and downstream response types.

[0025] Please refer to Figure 2 The acquisition steps of the behavior coordination quantification result are as follows: S101: Obtain the brain glioma tissue expression data, detect the expression arrangement order of the cell markers in the sample sequence, construct the expression segmentation interval according to the expression value in turn, calculate the expression mean change direction of each marker in the corresponding segmentation interval, and generate the segmentation expression change direction data; Obtaining brain glioma tissue expression data refers to downloading and standardizing the RNA-seq expression matrix from the TCGA or CGGA database, eliminating the sequencing depth difference of different samples by using the TPM conversion method, calling the expression data of the cell marker gene in the sample sequence, sorting the expression values from large to small to construct the sequence position, dividing the samples into multiple expression segmented intervals according to the expression value sorting results, for example, 60 samples are divided into 6 intervals according to every 10 samples, the average expression of the marker is calculated in each interval, and the average value change trend of each marker in the interval is recorded, if the average values of a marker in the first two intervals are 8.2 and 10.5 respectively, it is recorded as an upward trend, otherwise if they are 8.2 and 7.6, it is recorded as a downward trend, if the change value is within 0.3, it is recorded as no change, the value can be set as a change judgment threshold according to the average standard deviation of the marker in the sample group, and the segmented direction vector of each marker is constructed in this way, for example, the marker A forms 5 trend directions in 6 intervals, which are upward, downward, upward, no change and upward, and the direction value is recorded as {+1, -1, +1, 0, +1}, the direction vectors of multiple markers are stored in the marker expression trend structure table as a reference standard for subsequent candidate factor judgment.

[0026] S102: Based on the segmented expression change direction data, the expression average value change direction of the candidate factor in each segmented interval is calculated, the expression change direction of the candidate factor and each cell marker in the same segmented interval is compared, the consistent change direction in each interval is recorded, and the trend corresponding statistical value is obtained; Based on the segmented expression change direction data constructed above, the expression data of the candidate factor in the sample is called, and the candidate factor expression data is also divided into 6 intervals according to the same interval division method as the marker, the average value in each interval is calculated, and the average value comparison between adjacent intervals is carried out to generate the expression direction vector, for example, the average values of a certain candidate factor in 6 intervals are 3.2, 4.1, 4.0, 4.3, 3.9 and 4.2 respectively, and the expression direction is {+1, 0, +1, -1, +1}, then the direction vector of each candidate factor is compared with the direction vector of all markers, and it is judged whether the direction in each interval is consistent, if the direction of the candidate factor in the first interval is +1, and 3 out of 5 markers are +1, then the consistent trend number in this interval is 3, and the comparison process of all intervals is completed in turn, the number of times that each candidate factor is consistent with all markers in each interval is recorded, forming the trend consistent statistical matrix corresponding to the candidate factor and the interval, the row in the matrix represents the candidate factor, the column represents the segmented interval, and the cell records the consistent number, which is the basis for subsequent trend cooperativity quantification.

[0027] S103: Call the trend corresponding statistical value, count the number of times that the candidate factor and the cell marker change trend are consistent in each segmented interval, and combine the interval distribution rule to establish a numerical scale reflecting the overall expression behavior synergy of the candidate factor, and obtain the behavior synergy quantization result; Call the trend consistent statistical matrix, traverse the number of times that each candidate factor is consistent in all segmented intervals along the row direction, and count the number of intervals in which the direction is consistent in all intervals. Then, in combination with the theoretical maximum number of consistent times of the candidate factor, that is, the product of the number of markers participating in comparison and the number of intervals, calculate the interval coverage rate and the frequency normalization value, and take their average to constitute the synergy score. For example, if a factor is compared with 8 markers, there are 5 intervals in which the direction is consistent in 6 intervals, a total of 32 consistent times, and the theoretical maximum value is 8*6=48. Then the interval coverage rate is 5 / 6, the frequency normalization value is 32 / 48, and the synergy score is the average of the above two values. The synergy scores of all candidate factors are output in vector form as a quantitative indicator of the expression behavior synergy of the factors, and the behavior synergy quantization result is finally obtained.

[0028] Please refer to Figure 3 , the obtaining step of the trajectory splitting node structure section is: S201: Based on the behavior synergy quantization result, judge the expression dense section of the candidate factor in the continuous sample, detect the aggregation mode of the expression value in the sample arrangement, and screen the differential expression amplitude of the expression dense interval to generate the expression dense interval differential coefficient; Based on the behavior synergy quantization result, the expression value sequence of each candidate factor is called in turn, and the expression aggregation state formed by the candidate factor in the sample arrangement order is judged. First, arrange the samples in the original number order, and then calculate the local expression mean and standard deviation of the candidate factor in the equidistant sliding window to form a sliding expression statistical sequence. For example, assuming that the window size is 10 and the step size is 1, 51 sliding window expression mean and standard deviation values are formed for a factor in 60 samples. Then, identify the continuous window segment in which the expression mean is greater than the overall mean and the standard deviation is less than the overall standard deviation. The dense section identification condition is set as the expression mean being higher than 1.2 times the overall mean and the standard deviation being less than 0.8 times the overall standard deviation. The start and end sample positions in the continuous window segment that meets the condition are extracted as the expression dense section. Then, for all identified dense sections, the interval differential expression value is constructed based on the difference between the section means. Further, the absolute value of the mean difference between all dense sections is calculated and standardized, and the expression intensity difference of each candidate factor is recorded to establish an expression change intensity sequence. The number of section combinations that can be compared is calculated based on the total number of samples and the number of sliding windows, and finally the expression dense interval differential coefficient is generated.

[0029] S202: Constructing interval demarcation point identification condition according to expression dense interval difference coefficient, judging expression breakpoint position, and combining with the subtype distribution to which the samples on both sides of the expression breakpoint belong, calculating the proportion of subtype distribution on both sides, and generating demarcation point subtype distribution coefficient; According to the expression dense interval difference coefficient, the interval expression difference value vector of each candidate factor is called to judge whether the expression change between adjacent dense segments meets the breakpoint judgment condition. The expression breakpoint identification condition is set as the difference between the expression means of adjacent segments being greater than 1.5 times the standard deviation. If the expression difference value of the candidate factor between the 2nd and 3rd dense segments is 2.1, and the overall expression standard deviation of the candidate factor in this segment is 1.3, then the breakpoint condition is established. The sample index position corresponding to the breakpoint is recorded as the candidate breakpoint position. Then, based on the samples on both sides of the breakpoint, the proportion of the corresponding subtype label distribution is calculated. For example, among the 12 samples on the left side of the breakpoint, 8 are of subtype A, and among the 12 samples on the right side, 3 are of subtype A. The proportion of subtype A on both sides is 0.67 and 0.25, respectively. The absolute value of the difference between the subtype proportions is 0.42. Further, the process is repeated for all candidate breakpoints to form a subtype distribution difference value vector corresponding to the candidate breakpoint position, and the vector is normalized. The minimum difference threshold is set to 0.2, and the breakpoint positions below the threshold are deleted. The breakpoint positions and their corresponding subtype proportion differences that pass the screening are retained, and finally the demarcation point subtype distribution coefficient is generated.

[0030] S203: Calling demarcation point subtype distribution coefficient, performing group difference test, screening the positions where expression interval splitting occurs, and combining the splitting interval positioning with branch information to establish node structure value and obtain trajectory splitting node structure segment; According to the demarcation point subtype distribution coefficient, the group difference test is performed on the samples on both sides of each candidate breakpoint. The expression value is used as the basis for the test. The mean and variance of the candidate factor in the samples on both sides of the breakpoint are calculated. The appropriate variance homogeneity test method is used for preprocessing according to the sample size. The significance threshold is set to 0.05. If the p-value of the difference test is less than the threshold, it is determined that there is expression splitting behavior at this position. The sample index position is recorded as the effective splitting breakpoint position. Then, the position is mapped to the specific number in the sample order. The original expression distribution structure of the candidate factor is called. The expression rising or falling trend before and after the splitting position and the subtype label conversion trend are combined to construct the node trend structure. The node structure includes three attributes: sample position, expression change direction, and subtype change direction. Then, it is judged whether the candidate factor has branch behavior according to whether there are multiple effective splitting positions. If there are 2 or more effective breakpoints, and the subtypes to which the samples corresponding to the breakpoints belong show a jump-like conversion, it is recorded that the candidate factor has a multi-branch structure. Finally, all effective splitting positions and their branch attributes are combined to form the trajectory splitting node structure segment.

[0031] Please refer toFigure 4 The specific steps for obtaining the functional pathway drift trend score are as follows: S301: Based on the structural segments of the trajectory split nodes, obtain the enrichment expression level of each functional pathway in multiple consecutive segment intervals of the immune score, analyze the changes in the enrichment level of each pathway in the score interval, and obtain the mean enrichment data of the pathway interval. Based on the structural segments of the trajectory split nodes, the enriched expression levels of each functional pathway in different scoring segments are extracted. First, using the scoring interval as an index, multiple continuous segmented intervals are constructed within the split structural segments. For example, the immune score is divided into five segments from lowest to highest, with each segment corresponding to several sample numbers. Expression matrix data of all samples in each segment are obtained. Genes associated with each functional pathway are aggregated, and the enriched expression level of that pathway in that segment is calculated using a weighted average method. The weights of the pathways are based on the KEGG or [other data / data / indices]. The importance of annotations in the GO database is determined. For example, if a pathway contains 5 genes with weights of 0.3, 0.25, 0.2, 0.15, and 0.1 respectively, the expression values ​​of the corresponding genes in that segment are multiplied by their weights and summed to obtain the enriched expression level of the pathway. After extracting the pathway expression in the 5 scoring intervals, the enrichment level of each functional pathway in the 5 segments is used to form a vector and recorded as the interval expression trajectory. The above operation is repeated for all pathways to finally obtain the mean expression sequence of each pathway in all scoring intervals, forming the mean enrichment data of the pathway intervals.

[0032] S302: Call the mean enrichment data of the channel interval, compare the change of enrichment direction of each functional channel in the scoring interval, filter the channel sequence with consistent enrichment direction in the scoring interval, count the continuous change interval of the channel, and obtain the value of the continuous change interval of the channel. The enrichment mean data of each pathway is retrieved, and the enrichment expression change direction of each functional pathway in different scoring segments is compared sequentially. If the enrichment mean of a pathway in scoring segments 1 to 5 is 1.2, 1.5, 1.7, 2.0, and 2.3 respectively, then its direction is continuously rising. If it is 2.0, 1.9, 2.1, 2.0, and 1.8 respectively, then the direction is inconsistent. The consistency judgment criterion for enrichment direction is that the enrichment value changes in the same direction in three consecutive segments. For example, if all three consecutive scoring segments are rising or falling, then it is judged as consistent direction. The consecutive scoring segments that meet this condition are extracted and recorded as consistent pathway segments. After performing this judgment process on all pathways, the number of pathways and segments that meet the consistent direction condition are counted. For example, if pathway A rises continuously in scoring segments 1 to 4, then the value of the continuous change interval is 3. A statistical table of the number of consistent direction intervals is established for all pathways, and finally the value of the continuous change interval of the pathway is obtained.

[0033] S303: Based on the pathway continuous change interval value, the continuity and interval change amplitude of the enrichment trend of each functional pathway in the continuous scoring interval are analyzed, a pathway activity dynamic trajectory index is constructed, and a functional pathway drift trend score is generated; Based on the pathway continuous change interval value, the enrichment trend stability of each functional pathway in the continuous scoring section is calculated. First, the expression mean variance in each pathway continuous section is calculated as the trend stability index. The stability benchmark is set as the expression value standard deviation in the section being less than 0.7 times the standard deviation of the whole scoring interval. If the expression mean of pathway B in three continuous sections is 1.4, 1.5, and 1.6, and the standard deviation is 0.1, while the standard deviation of pathway B expression in the whole scoring interval is 0.2, then the stability requirement is met. In addition, the enrichment value section span of each pathway is calculated, and the difference between the highest interval expression value and the lowest interval expression value is calculated as the change amplitude. The trend stability index and the change amplitude are combined to construct the activity dynamic trajectory score of the pathway. The scoring formula can be set as a weighted average of the two, with the stability weight being 0.6 and the change amplitude weight being 0.4. When the stability score of pathway B is 0.9 and the change amplitude score is 0.7, the activity dynamic score is 0.82. The index of all pathways is calculated in turn to finally form the functional pathway drift trend score.

[0034] Please refer to Figure 5 The acquisition steps of the regulatory core evaluation result are as follows: S401: According to the functional pathway drift trend score, the output data of each candidate factor under the three labels of immune score, prognosis score, and subtype classification is obtained, the contribution difference value of each label output is analyzed, the contribution difference interval of each factor is calculated, and the label contribution interval value is obtained; According to the functional pathway drift trend score, the output data of each candidate factor under the three labels of immune score, prognosis score, and subtype classification is extracted, and the scoring results are normalized to the interval of 0 to 1, which is convenient for subsequent difference calculation. The original scoring data can be processed by Z-score standardization to eliminate the scale difference between data. For example, the candidate factor X is 0.85 in immune score, 0.73 in prognosis score, and 0.91 in subtype classification. The output difference value of the factor between each two labels is calculated in turn, i.e. the difference value between immune and prognosis is 0.12, the difference value between immune and subtype is 0.06, and the difference value between prognosis and subtype is 0.18. The maximum and minimum values of the three difference values are judged and the interval in which they are located is recorded. For example, the interval is divided into 0-0.1, 0.1-0.2, etc. with a step of 0.1. If the maximum difference value of factor X is 0.18, the contribution difference interval of factor X is 0.1-0.2. The maximum difference value of all factors in the three label outputs is divided into the corresponding interval, the number of factors in each interval is counted, and a factor contribution difference interval distribution table is established to finally obtain the label contribution interval value.

[0035] S402: Based on the label contribution interval value, compare the contribution difference range between the label output, judge the stability of the difference degree and adjust the weight of the corresponding label, and obtain the label output processing result; Based on the label contribution interval value, the contribution difference range of each candidate factor between different labels is extracted, and whether the fluctuation amplitude of the scoring result is stable is judged accordingly. The difference stability judgment threshold is set to 0.15, that is, if the difference between the maximum score and the minimum score in three labels is less than or equal to 0.15, the factor is less volatile in label output, and is considered stable, if it exceeds the value, it is larger, the number of each type of factor is counted, and the proportion of stable factors is obtained. Then, according to the proportion, the label weight is adjusted, for example, when most of the factors have the smallest fluctuation in the prognosis score, the prognosis score weight is increased to 0.5, the immune and subtype scores are set to 0.3 and 0.2 respectively, and all the weights are added to 1. Set the weight configuration, recalculate the score of each factor, for example, the original scores of candidate factor Y in three labels are 0.6, 0.8 and 0.7 respectively, the new score is 0.6*0.3 + 0.8*0.5 + 0.7*0.2 = 0.71, and the score result of each candidate factor is recorded as the label output processing result after executing the calculation process.

[0036] S403: Call the label output processing result, summarize the output of each candidate factor in all label combinations after weighting, evaluate the cross-label stable influence degree, establish the factor stable influence score, identify and screen the key factor, and obtain the core regulatory evaluation result; Call the label output processing result, and calculate the weighted scores of each candidate factor in three labels to evaluate the stable influence degree of the factor across labels. The stable influence evaluation standard is set to a coefficient of variation CV (standard deviation divided by mean) less than 0.2. If the condition is met, it indicates that the scores of the factor in different labels are close, and the influence is stable. The CV value is calculated by taking the standard deviation and mean of the weighted scores of each factor in three labels, for example, the weighted scores of candidate factor Z in three labels are 0.72, 0.69 and 0.71, the mean is 0.7067, the standard deviation is 0.0153, and the CV is about 0.0217, which is less than the stable evaluation standard. Therefore, the factor is evaluated as a stable influence factor. Sort the CV values of all factors, and select the top 20% of factors with the lowest CV values as key factors. Summarize all label weighted scores and stability evaluation results to form tabular record data, and finally obtain the core regulatory evaluation result.

[0037] Please refer to Figure 6 The acquisition step of the direction path semantic ranking information is specifically: S501: According to the evaluation results of the regulatory core, the arrangement order of the candidate factor name and the brain glioma related term in the sentence is detected in multiple relevant literature paragraphs, the relative order of each group of words is recorded, the order statistical table of the factor in the literature is established, and the factor word order statistical data is obtained; According to the evaluation results of the regulatory core, the brain glioma related research articles collected in the literature database are retrieved one by one, the sentence structure in which the candidate factor name and the brain glioma core term (such as "proliferation", "infiltration", "immunosuppression", "angiogenesis", etc.) coexist is screened, the position relationship of each candidate factor name and the brain glioma term in the sentence is recorded as a pair of word order items, for example, the sentence expression in a certain literature is "NF-κB activation leads to glioma cell infiltration", which is recorded as "NF-κB→infiltration", and in another sentence such as "IL-6 plays an important role in the development of glioma", it is recorded as "IL-6←glioma". The direction vector of the arrangement of each group of word pairs in the sentence is constructed in this way, the candidate factors appearing in all literature paragraphs are traversed, and the relative order direction between each factor and the brain glioma term is counted, the number of forward or backward is recorded, and the order statistical frequency matrix is formed. After standardizing the word direction distribution of each candidate factor in the matrix, the word order statistical data of the factor in the literature is summarized and established.

[0038] S502: Based on the factor word order statistical data, the number of times of the candidate factor before and after the brain glioma related term in multiple literature paragraphs is analyzed, and the word order distribution index is calculated; Based on the word order statistical data, the position relationship of the candidate factor name before and after the brain glioma related term in the literature paragraph is classified and marked, and the frequency of each candidate factor appearing before (leading) the term and appearing after (postfix) the term is counted, and the proportion of the word order bias is expressed, that is, the word order distribution index calculation formula is: word order distribution value= , for example, IL-6 coexists with the brain glioma term for 80 times in 100 literature paragraphs, of which 56 times are leading and 24 times are postfixed, and the word order distribution value is , that is, the word order is biased towards the leading type. The word order distribution ratio of each factor is obtained by such calculation, and the index is used to measure whether it is more likely to be a starting regulatory factor or a response factor in the literature expression, a word order dominant type classification table of the candidate factor is established, and the word order distribution index is finally obtained.

[0039] S503: The position of the target factor in the regulation chain is judged by calling the word order distribution index, including the starting regulator and the response participant, the semantic label and the word order distribution index of each candidate factor are summarized, the semantic direction label is established, and the direction path semantic sorting information is obtained; The call order distribution index is used to determine the role according to the distribution ratio of each factor in the literature order, the threshold is set to 0.6, if the order distribution value is greater than 0.6, it is determined that it is more biased as the starting regulator, if it is less than 0.4, it is a response participant, if it is between 0.4 and 0.6, it is classified as a neutral regulatory associated person, the above determination rule is bound with the order distribution value, and the original literature is labeled with co-occurrence semantic keywords, such as "induction", "start", which further confirms the identity of the regulator, "mediation", "response" is marked as a participant identity, finally the semantic role label of each factor is generated, and the role category and the order distribution value are matched and arranged, and are summarized into a semantic positioning table, each record contains the factor name, the order value, the role category and the semantic keyword, which is used to describe the semantic directionality of the factor in the regulation chain, and finally the directional path semantic order information is established.

[0040] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for analyzing the key molecular mechanism of brain glioma microenvironment formation, characterized in that, Includes the following steps: S1: Call the expression data of glioma tissue, analyze the expression order of cell markers in the sample sequence, construct segmented intervals, calculate the expression change direction of candidate factors in each interval, combine the expression trend of markers to make direction correspondence judgment, count the number of trend consistency, and construct behavioral synergy quantification results; S2: Using the behavioral collaborative quantification results, determine the expression clustering of candidate factors in continuous samples, combine expression differences and subtype distribution differences, identify expression splitting locations and construct branch structure markers, and generate trajectory splitting node structure segments; S3: Using the aforementioned trajectory splitting node structural segments, compare the expression levels of functional pathways within the immune score segment intervals, analyze pathway expression trends and interval changes, construct a dynamic trajectory index for pathway activation, and generate a functional pathway drift trend score. S4: Using the functional pathway drift trend score, compare the output differences of candidate factors under multiple labels, compare the contribution difference range between label outputs, determine the stability of the degree of difference and adjust the weight of the corresponding label, evaluate the degree of cross-label stability influence, identify key factors, and generate the core evaluation results of regulation. 2.The method of claim 1, wherein the method is characterized by, The behavioral coordination quantification results include the number of times the direction corresponds, the trend coverage, and the sequence of expression change directions. The trajectory split node structure segment specifically includes the split region positioning information, subtype distribution comparison records, and node branch identifiers. The functional pathway drift trend score includes the pathway continuous change trend, the score interval span index, and the enrichment direction change characteristics. The regulation core evaluation results specifically include the cross-label contribution score, the label stability coefficient, and the key factor screening number. 3.The method of claim 1, wherein the method is characterized by, The constructed pathway activation dynamic trajectory index refers to the recording of the expression enhancement trend and the number of sustained intervals of the pathway in each interval based on the changes in the enrichment level of functional pathways in multiple immune score segments, reflecting the dynamic characteristics of the pathway response under various immune states, and measuring the degree of activity in microenvironment regulation. 4.The method of claim 1, wherein the method is characterized by, The specific steps for obtaining the behavioral collaboration quantification results are as follows: S101: Obtain expression data of glioma tissue, detect the expression order of cell markers in the sample sequence, construct expression segment intervals according to the expression level values, calculate the direction of change of the mean expression of each marker in the corresponding segment interval, and generate segment expression change direction data. S102: Based on the segmented expression change direction data, calculate the average expression change direction of the candidate factor in each segment interval, compare the expression change direction of the candidate factor with that of each cell marker in the same segment interval, record the cases where the change direction is consistent in each interval, and obtain the trend corresponding statistical value. S103: Call the statistical values ​​corresponding to the trend, count the number of times the candidate factor and cell marker change trends are consistent in each segment interval, and combine the interval distribution pattern to establish a numerical scale reflecting the overall expression behavior synergy of the candidate factor, and obtain the behavioral synergy quantification results. 5.The method of claim 4, wherein the method is characterized by, The specific steps for obtaining the trajectory splitting node structure segment are as follows: S201: Based on the behavioral co-quantification results, determine the expression-dense regions of candidate factors in continuous samples, detect the clustering pattern of expression values ​​in the sample arrangement, screen the differential expression amplitude of expression-dense regions, and generate the differential coefficient of expression-dense regions. S202: Based on the expression density interval difference coefficient, construct interval boundary point identification conditions, determine the expression breakpoint location, and combine the subtype distribution to which the samples on both sides of the expression breakpoint belong, calculate the subtype distribution ratio on both sides, and generate the boundary point subtype distribution coefficient. S203: Call the subtype distribution coefficient of the dividing point, perform group difference test, screen the location where the expression interval splits, and combine the split interval location with the branch information to establish node structure values ​​and obtain the trajectory split node structure segment.

6. The method for elucidating the key molecular mechanism of glioma microenvironment formation according to claim 5, characterized in that, The specific steps for obtaining the functional pathway drift trend score are as follows: S301: Based on the trajectory split node structure segment, obtain the enrichment expression level of each functional pathway in multiple consecutive segment intervals of the immune score, analyze the change of the enrichment level of each pathway in the score interval, and obtain the mean enrichment data of the pathway interval. S302: Call the mean enrichment data of the pathway interval, compare the change of enrichment direction of each functional pathway in the scoring interval, filter the pathway sequence with consistent enrichment direction in the scoring interval, count the continuous change interval of the pathway, and obtain the value of the continuous change interval of the pathway. S303: Based on the numerical values ​​of the continuous change interval of the pathway, analyze the persistence of the enrichment trend of each functional pathway in the continuous scoring interval and the range of interval change, construct the pathway activation dynamic trajectory index, and generate functional pathway drift trend score.

7. The method for elucidating the key molecular mechanism of glioma microenvironment formation according to claim 6, characterized in that, The specific steps for obtaining the core assessment results of the regulation are as follows: S401: Based on the functional pathway drift trend score, obtain the output data of each candidate factor under the three labels of immune score, prognostic score and subtype classification, analyze the contribution difference of each label output, and calculate the contribution difference interval of each factor to obtain the label contribution interval value. S402: Based on the value of the label contribution interval, compare the contribution difference range between label outputs, determine the stability of the difference degree and adjust the weight of the corresponding label to obtain the label output processing result; S403: Call the label output processing results, summarize the output of each candidate factor after weighting in all label combinations, evaluate the degree of cross-label stable influence, establish factor stable influence score, identify and screen key factors, and obtain the core evaluation results of regulation.

8. The method for elucidating the key molecular mechanism of glioma microenvironment formation according to claim 1, characterized in that, The method further includes: S5: Using the evaluation results of the aforementioned regulatory core, detect the word order structure of candidate factors and glioma terms in multiple articles, analyze the word order position distribution, establish semantic directional labels, and generate directional path semantic ranking information; The directional path semantic sorting information includes semantic preorder occurrence frequency, control direction distribution labels, and path structure positioning index.

9. The method for elucidating the key molecular mechanism of glioma microenvironment formation according to claim 8, characterized in that, The establishment of semantic directional labels refers to the process of determining the semantic position of candidate factors in the regulatory pathway based on the sentence order of candidate factors and glioma-related terms in the literature, counting the frequency ratio of terms appearing at the beginning or end of the expression, and classifying them into upstream regulation and downstream response types based on directional preferences.

10. The method for elucidating the key molecular mechanism of glioma microenvironment formation according to claim 8, characterized in that, The specific steps for obtaining the directional path semantic ordering information are as follows: S501: Based on the evaluation results of the regulatory core, detect the order of candidate factor names and glioma-related terms in sentences in multiple relevant literature paragraphs, record the relative order of each group of words, establish a statistical table of factor order in the literature, and obtain factor word order statistics. S502: Based on the statistical data of the factor word order, analyze the proportion of times the candidate factor appears before and after terms related to glioma in multiple literature paragraphs, and calculate the word order distribution index. S503: Call the word order distribution index to determine the position of the target factor in the regulatory chain, including the initiator and the responder. Summarize the semantic labels and word order distribution index of each candidate factor, establish semantic directional labels, and obtain directional path semantic ranking information.

Images