Sample information tracing method for kidney disease biological sample library

Abstract

The invention relates to the technical field of abnormal process link prediction, in particular to a sample information tracing method for a kidney disease biological sample library. The method comprises the following steps: acquiring an abnormal kidney disease biological sample circulation process and constructing an FP tree; for any support degree threshold value in a preset support degree threshold value set, processing the FP tree through the support degree threshold value to obtain a maximum frequent item set; obtaining the reference weight of the maximum frequent item set based on the value and depth of each node of the path corresponding to the maximum frequent item set in the FP tree, and further obtaining the reasonable degree of the support degree threshold; and obtaining an optimal support degree threshold based on the reasonable degree, and predicting an abnormal process link of the kidney disease biological sample according to the optimal support degree threshold. According to the method, the optimal support degree threshold is obtained, so that the circulation process which easily causes the abnormality of the kidney disease biological sample is accurately predicted, the abnormal risk is processed and avoided in time, and the abnormal phenomenon of the kidney disease biological sample is effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of predicting abnormal process links, and in particular to a method for tracing sample information of a kidney disease biological sample library. Background Art

[0002] Kidney disease is a disease that affects kidney function and structure. In order to develop effective drugs for kidney disease, it is necessary to study kidney disease biological samples. In the process of studying kidney disease biological samples, it is necessary to ensure the effectiveness and normality of kidney disease biological samples. In actual situations, kidney disease biological samples are prone to abnormal phenomena such as loss and contamination during the process from collection to research. In order to avoid abnormal kidney disease biological samples, it is necessary to trace the entire process of kidney disease biological samples from collection, distribution, use and destruction, analyze the abnormalities of specific links and processes, increase risk control in a timely manner, and effectively avoid the causes of abnormal kidney disease biological samples.

[0003] In the existing method, the process information of abnormal kidney disease biological samples is traced through the FP tree (Frequent Pattern Tree), and the abnormal link process is predicted according to the artificially set support threshold, so as to timely handle and avoid risks of the abnormal circulation process of kidney disease biological samples, and reduce the abnormal phenomenon of kidney disease biological samples. However, in actual situations, the artificially set support threshold is inaccurate. In the process of tracing the process information of abnormal kidney disease biological samples through the FP tree (Frequent Pattern Tree), too large or too small support threshold will lead to inaccurate prediction of abnormal link process, resulting in the inability to accurately predict the circulation process that is likely to cause abnormal kidney disease biological samples, and thus the inability to accurately and timely handle and avoid abnormal risks, and the inability to effectively reduce the abnormal phenomenon of kidney disease biological samples. Summary of the invention

[0004] In order to solve the technical problem that the artificially set support threshold is inaccurate, resulting in inaccurate prediction of abnormal link processes, and thus failing to effectively reduce the abnormality of kidney disease biological samples, the purpose of the present invention is to provide a sample information traceability method for a kidney disease biological sample library. The technical solution adopted is as follows: An embodiment of the present invention provides a method for tracing the sample information of a kidney disease biological sample library, the method comprising the following steps: Obtain the flow process of each abnormal kidney disease biological sample and construct an FP tree; for any support threshold in the preset support threshold set, process the FP tree through the support threshold to obtain the maximum frequent item set; Based on the value and depth of each node in the FP tree corresponding to each maximum frequent item set, obtain the reference weight of each maximum frequent item set; According to the reference weight, obtaining the rationality of the support threshold; The optimal support threshold is obtained based on the reasonable degree, and the abnormal process links of kidney disease biological samples are predicted according to the optimal support threshold.

[0005] Furthermore, the reference weight is obtained by: For any maximum frequent item set, the path corresponding to the maximum frequent item set is used as the target path, and the nodes in the target path are divided according to the value size to obtain the first category and the second category; wherein the value of the node in the first category is larger, and the value of the node in the second category is smaller; According to the numerical difference between the nodes in the first category and the nodes in the second category, the similarity of the nodes on the target path is obtained; The node with the largest value in the second category is used as the split node of the target path, and the first abnormality possibility degree of the first category is obtained according to the numerical difference between the split node and other nodes at the same depth in the FP tree, and the number of all nodes at the depth of the split node in the FP tree; The node with the smallest value in the second category is taken as the end point of the target path, and the second abnormal possibility degree of the second category is obtained according to the fluctuation of the value of the node in the second category and the similarity between the value of the node at the next depth of the end point in the FP tree and the value of the node in the second category; According to the similarity degree, the first abnormal possibility degree and the second abnormal possibility degree, obtaining the overall abnormality degree of the maximum frequent item set; Obtaining the importance of the maximum frequent itemset according to the depth of the end point in the FP tree and the value of each node in the target path; The product of the overall abnormality degree and the importance degree is used as the reference weight of the maximum frequent item set.

[0006] Furthermore, the method for obtaining the similarity degree is: Obtaining the average value of the values ​​of all nodes in the first category as a first reference value; Obtaining the average value of the values ​​of all nodes in the second category as a second reference value; The result of negatively correlating and normalizing the difference between the first reference value and the second reference value is used as the similarity of the nodes on the target path.

[0007] Furthermore, the method for obtaining the first abnormality possibility degree is: The depth of the split node in the FP tree is taken as the target depth, and the average value of the values ​​of all nodes except the split node at the target depth is obtained as the first eigenvalue; The difference between the value of the segmentation node and the first characteristic value is used as a reference abnormality degree of the first category; The result of normalizing the product of the inverse of the number of all nodes at the target depth and the reference abnormality degree is taken as the first abnormality possibility degree of the first category.

[0008] Furthermore, the method for obtaining the second abnormality possibility degree is: The result of negatively correlating and normalizing the standard deviation of the values ​​of the nodes in the second category is used as the initial abnormal possibility degree of the second category; Obtain the variance of the values ​​of the nodes in the second category as the first variance; Obtaining the variance of the values ​​of all nodes at the next depth of the end point in the FP tree and the values ​​of the nodes in the second category as the second variance; The result of normalizing the absolute value of the difference between the first variance and the second variance is used as the correction weight of the second category; The product of the revised weight of the second category and the initial abnormality possibility degree is taken as the second abnormality possibility degree of the second category.

[0009] Furthermore, the method for obtaining the overall abnormality degree is: Obtaining the average of the first abnormality possibility degree and the second abnormality possibility degree as the overall abnormality possibility degree of the maximum frequent item set; The product of the similarity degree and the overall abnormality possibility degree is taken as the overall abnormality degree of the maximum frequent item set.

[0010] Furthermore, the method for obtaining the importance is: The ratio of the average value of all nodes in the target path to the depth of the end point in the FP tree is used as the importance of the maximum frequent itemset.

[0011] Furthermore, the method for obtaining the reasonable degree is: The mean of the reference weights of all maximum frequent itemsets is taken as the reasonableness of the support threshold.

[0012] Furthermore, the method for obtaining the optimal support threshold is: The support threshold corresponding to the maximum reasonable degree is taken as the optimal support threshold.

[0013] Furthermore, the method for obtaining the abnormal process link is: The FP tree is processed by the optimal support threshold, and each maximum frequent item set obtained at this time is used as the target item set; Obtain the flow process of each normal kidney disease biological sample and construct an FP tree as a reference FP tree, process the reference FP tree through the optimal support threshold, and use each maximum frequent item set obtained at this time as a reference item set; Each target item set is compared with each reference item set, and the process links of the paths corresponding to the target item sets that are different from the reference item sets are regarded as abnormal process links of the kidney disease biological samples.

[0014] The present invention has the following beneficial effects: The present invention constructs an FP tree according to the flow of abnormal kidney disease biological samples, avoids the interference of normal kidney disease biological samples, and is conducive to efficiently analyzing the process that causes abnormal kidney disease biological samples; in order to obtain a reasonable support threshold, the process link that causes abnormal kidney disease biological samples can be accurately predicted, and then for any support threshold in the preset support threshold set, the FP tree is processed by the support threshold to obtain the maximum frequent item set, avoiding the analysis of repeated process links; in order to accurately analyze whether the support threshold is reasonable, then based on the value and depth of each node of the path corresponding to each maximum frequent item set in the FP tree, the maximum frequent item set is obtained. The reference weight of each maximum frequent item set accurately reflects the possibility that the process link corresponding to each maximum frequent item set causes abnormal kidney disease biological samples, which is conducive to the subsequent accurate analysis of whether the support threshold is reasonable; therefore, the reasonable degree of the support threshold is accurately obtained according to the reference weight, and the rationality of the support threshold is accurately reflected; then the optimal support threshold is obtained based on the reasonable degree, so that the process link that causes abnormal kidney disease biological samples is more accurate, and then the abnormal process link of kidney disease biological samples is accurately predicted according to the optimal support threshold, which is conducive to timely handling and avoidance of abnormal risks and effectively reducing the occurrence of abnormal phenomena in kidney disease biological samples. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 A schematic flow chart of a method for tracing the sample information of a kidney disease biological sample library provided by an embodiment of the present invention; Figure 2 A flow chart of a method for obtaining a reference weight provided by an embodiment of the present invention; Figure 3A structural diagram of a sample information traceability system for a kidney disease biological sample library provided by an embodiment of the present invention; Figure 4 A schematic diagram of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the sample information traceability method of a kidney disease biological sample library proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0018] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0019] The following is a detailed description of a method for tracing the sample information of a kidney disease biological sample library provided by the present invention in conjunction with the accompanying drawings.

[0020] Embodiment 1: This invention proposes a method for tracing the sample information of a kidney disease biological sample library. Figure 1 , which shows a schematic flow chart of a method for tracing sample information of a kidney disease biological sample library provided by an embodiment of the present invention, the method comprising the following steps: Step S1: Obtain the flow process of each abnormal kidney disease biological sample and construct an FP tree; for any support threshold in the preset support threshold set, process the FP tree through the support threshold to obtain the maximum frequent item set.

[0021] Specifically, kidney disease biological samples are first collected in the hospital, then transported to a dedicated storage center, and finally uniformly distributed to various research institutes or research institutes. The entire transportation process includes multiple process links. The storage center needs to first perform abnormality detection on kidney disease biological samples to ensure that the kidney disease biological samples distributed to the research institute or research institute are normal and ensure the significance of the research.

[0022] Because kidney disease biological samples may be damaged and abnormal during transportation due to environmental factors such as temperature and humidity, as well as unreasonable transportation methods, this embodiment selects a batch of kidney disease biological samples, traces the flow process of each kidney disease biological sample in the batch of kidney disease biological samples, and performs abnormality detection on each kidney disease biological sample in the batch of kidney disease biological samples after arriving at the storage center, and marks kidney disease biological samples with obvious DNA or protein abnormalities as abnormal kidney disease biological samples. In this embodiment, the number of kidney disease biological samples in the selected batch of kidney disease biological samples is set to 1000. The implementer can set the number of kidney disease biological samples in the batch according to actual conditions, and it is not limited here. Among them, each kidney disease biological sample is assigned a unique identifier, that is, a unique ID.

[0023] In order to analyze the process links that are likely to cause abnormal kidney disease biological samples, and taking into account the associations between different processes, this embodiment first obtains abnormal kidney disease biological samples, and then constructs an FP tree through the FP algorithm (Frequent Pattern Mining) according to each process of the flow of each abnormal kidney disease biological sample, and then analyzes the FP tree to predict the process links that are likely to cause abnormal kidney disease biological samples. It should be noted that the process links that are likely to cause abnormal kidney disease biological samples are composed of multiple related processes. Among them, the FP algorithm (Frequent Pattern Mining) is a well-known technology and will not be traced back.

[0024] By setting a support threshold for the FP tree to obtain frequent item sets, it is actually pruning the FP tree. Different support thresholds represent pruning at different positions of the FP tree. Therefore, different support thresholds correspond to different frequent item sets, and judging the correlation between different processes relies on frequent item sets to further predict the process links that are likely to cause abnormalities in biological samples of kidney disease. It is known that a support threshold that is too large or too small will lead to inaccurate predictions of process links that are likely to cause abnormalities in biological samples of kidney disease. Therefore, this embodiment first sets a preset support threshold set, and then analyzes the most reasonable support threshold in the preset support threshold set, and processes the FP tree to more accurately predict the process links that are likely to cause abnormalities in biological samples of kidney disease. It is known that the value range of the support threshold is 0 to 1. In order to improve the efficiency of obtaining the most reasonable support threshold, this embodiment sets the preset support threshold set to , implementers can set the preset support threshold set according to actual conditions, which is not limited here.

[0025] In order to analyze clearly, this embodiment analyzes each support threshold in the preset support threshold set separately, that is, for any support threshold in the preset support threshold set, the FP tree is processed by the support threshold to obtain the maximum frequent item set corresponding to the support threshold. The purpose of obtaining the maximum frequent item set in this embodiment is to avoid repeated analysis of the process associated with abnormal kidney disease biological samples. Therefore, this embodiment analyzes each maximum frequent item set. Among them, the method for obtaining the maximum frequent item set is a well-known technology and will not be repeated. It should be noted that the maximum frequent item sets analyzed subsequently are all corresponding to the support threshold.

[0026] Step S2: Based on the value and depth of each node in the FP-tree of the path corresponding to each maximum frequent item set, obtain the reference weight of each maximum frequent item set.

[0027] It is known that in the FP tree, a leaf node indicates that a specific item only appears with other items in the path corresponding to the frequent item set in the frequent item set. Therefore, the corresponding item of the leaf node, i.e., the process, has a low frequency of appearance in the process of abnormal kidney disease biological samples. When the FP tree is pruned by the set support threshold, the leaf node will be pruned first, which indirectly reflects that the process corresponding to the node with a greater depth, i.e., the farther away from the root node of the FP tree, is less likely to have no correlation with other processes, and the corresponding process appears less frequently in the process of abnormal kidney disease biological samples, which indirectly reflects that the corresponding process is less likely to cause abnormal kidney disease biological samples. Therefore, when the path corresponding to a certain maximum frequent item set has a greater depth in the FP tree, i.e., the farther away from the root node of the FP tree, the fewer times the process link of the path corresponding to the maximum frequent item set appears in the process link of abnormal kidney disease biological samples, and the process link of the path corresponding to the maximum frequent item set is less likely to cause abnormal kidney disease biological samples.

[0028] Considering that the larger the value of the node in the FP tree, the more times the process corresponding to the node appears in the process of abnormal kidney disease biological samples, which indirectly reflects that the process corresponding to the node is more likely to cause abnormal kidney disease biological samples. Therefore, when the values ​​of the nodes in the FP tree corresponding to the path of a certain maximum frequent item set are larger, the more times the process link of the path corresponding to the maximum frequent item set appears in the process link of abnormal kidney disease biological samples, the greater the possibility that the process link of the path corresponding to the maximum frequent item set causes abnormal kidney disease biological samples.

[0029] Therefore, this embodiment obtains the reference weight of each maximum frequent item set based on the value and depth of each node in the FP tree corresponding to the path of each maximum frequent item set. The larger the reference weight, the more likely the process link corresponding to the path corresponding to the maximum frequent item set is to be the process link that causes the abnormality of the kidney disease biological sample.

[0030] Preferably, in one possible implementation of this embodiment, the method for obtaining the reference weight is as follows: Figure 2 , which shows a flow chart of a method for obtaining a reference weight provided in this embodiment, the method comprising the following steps: Step S201: For any maximum frequent itemset, the path corresponding to the maximum frequent itemset is taken as the target path, and the nodes in the target path are divided according to the numerical values ​​to obtain a first category and a second category; wherein the numerical values ​​of the nodes in the first category are larger, and the numerical values ​​of the nodes in the second category are smaller.

[0031] In this embodiment, the K value in the K-means clustering algorithm is set to 2, and the nodes in the target path are divided according to the numerical values ​​of the nodes through the K-means clustering algorithm to obtain the first category and the second category. Among them, the K-means clustering algorithm is a well-known technology and will not be described in detail. Among them, the numerical values ​​of the nodes in the first category are relatively large, indicating that the nodes in the first category are shared with high frequency in the path, that is, the process corresponding to the nodes in the first category appears frequently in the process passed by the abnormal kidney disease biological sample; the numerical values ​​of the nodes in the second category are relatively small, indicating that the process corresponding to the nodes in the second category appears at a low frequency in the process passed by the abnormal kidney disease biological sample.

[0032] It should be noted that if the numerical values ​​of all nodes in the target path are the same, the nodes in the target path are classified according to the depth of the nodes in the target path by using the K-means clustering algorithm to obtain the first category and the second category. Among them, the depth of the nodes in the first category is smaller, and the depth of the nodes in the second category is larger.

[0033] Step S202: obtaining the similarity of nodes on the target path according to the numerical difference between the nodes in the first category and the nodes in the second category.

[0034] When the values ​​of the nodes on the target path are more equal, it means that the correlation between the processes in the maximum frequent item set corresponding to the target path is stronger, which indirectly indicates that the process link corresponding to the target path is a process link that abnormal kidney disease biological samples frequently pass through; when the numerical difference of the nodes on the target path is larger, it indirectly indicates that the process link corresponding to the target path contains some process links that abnormal kidney disease biological samples do not pass through, that is, it is not a process link that abnormal kidney disease biological samples frequently pass through. In the process link that predicts the abnormality of the kidney disease biological sample, the process link corresponding to the target path is less likely to cause the kidney disease biological sample to be abnormal. Therefore, this embodiment obtains the similarity of the nodes on the target path according to the numerical difference between the nodes in the first category and the second category. The greater the similarity, the greater the possibility that the process link corresponding to the target path causes the kidney disease biological sample to be abnormal.

[0035] Preferably, in a method that can be implemented in this embodiment, the method for obtaining the degree of similarity is: obtaining the mean of the values ​​of all nodes in the first category as the first reference value; obtaining the mean of the values ​​of all nodes in the second category as the second reference value; when the first reference value is more equal to the second reference value, it means that the values ​​of the nodes on the target path are more equal, and the corresponding process of the nodes on the target path is more likely to be the process that abnormal kidney disease biological samples pass through, which indirectly indicates that the process link corresponding to the target path is more likely to cause abnormal kidney disease biological samples. Then, the difference between the first reference value and the second reference value is obtained, and the result of negative correlation and normalization is used as the degree of similarity of the nodes on the target path. Among them, the difference between the first reference value and the second reference value must be greater than 0.

[0036] The calculation formula for similarity is: ; Where S is the similarity of nodes on the target path; is the first reference value; is the second reference value; exp is an exponential function with a natural constant as the base.

[0037] Step S203: Use the node with the largest value in the second category as the split node of the target path, and obtain the first abnormality possibility degree of the first category according to the numerical difference between the split node and other nodes at the same depth in the FP tree, and the number of all nodes at the depth of the split node in the FP tree.

[0038] In actual situations, temperature, humidity, sealing and other issues may cause abnormalities in kidney disease biological samples in different process links. Therefore, when the similarity of the nodes on the target path is small, some of the process links corresponding to the target path may also cause abnormalities in kidney disease biological samples, which indirectly indicates that the process links corresponding to the target path may cause abnormalities in kidney disease biological samples. It is known that the values ​​of the nodes change in order from small to large according to the path depth. Therefore, the first category and the second category in step S201 are actually segmenting the target path. In order to more accurately analyze whether the process links corresponding to the target path have an impact on the abnormality of kidney disease biological samples, this embodiment analyzes the first category and the second category respectively to determine the possibility that some of the process links corresponding to each stage of the target path will cause abnormalities in kidney disease biological samples.

[0039] In this embodiment, the node with the largest value in the second category is first used as the split node of the target path, that is, the target path is divided into two sections. When the value of the split node is larger, it means that the value of the node in the first category is larger, which indirectly reflects that the process link of the corresponding path of the first category is a process link that abnormal kidney disease biological samples frequently pass through together. Considering that if faced with abnormal kidney disease biological samples of different scales, there will be a large difference in the value of the node, so only analyzing the value of the node will have errors. However, when the value of the split node is greater than the value of other nodes at the same depth as the split node in the FP tree, and the fewer nodes at the depth of the split node in the FP tree, it can be accurately explained that the corresponding path of the first category has a higher node sharing situation than other paths, which indirectly reflects that the process link of the corresponding path of the first category is a process link that abnormal kidney disease biological samples frequently pass through together. Therefore, this embodiment obtains the first abnormal possibility of the first category according to the numerical difference between the split node and other nodes at the same depth in the FP tree, and the number of all nodes at the depth of the split node in the FP tree. The greater the possibility of the first abnormality, the greater the possibility that the process link of the path corresponding to the first category is a process link that abnormal kidney disease biological samples frequently pass through.

[0040] Preferably, in a method that can be implemented in this embodiment, the method for obtaining the first abnormal possibility degree is: taking the depth of the split node in the FP tree as the target depth, obtaining the mean of the values ​​of all nodes except the split node at the target depth as the first eigenvalue; taking the difference between the value of the split node and the first eigenvalue as the reference abnormality degree of the first category; the greater the reference abnormality degree, the higher the node sharing situation of the corresponding path of the first category compared with other paths. At the same time, when the number of all nodes at the target depth is smaller, it indirectly indicates that the corresponding path of the first category has a higher node sharing situation. Then, the result of normalizing the inverse of the number of all nodes at the target depth and the product of the reference abnormality degree is used as the first abnormal possibility degree of the first category.

[0041] The calculation formula for the first abnormal possibility degree is: ; In the formula, q is the first abnormal possibility of the first category; n is the number of all nodes at the target depth; E is the value of the split node; is the first eigenvalue; is the reference abnormality degree; norm is the normalization function.

[0042] Step S204: Take the node with the smallest value in the second category as the end point of the target path, and obtain the second abnormality possibility degree of the second category according to the fluctuation of the value of the node in the second category and the similarity between the value of the node at the next depth of the end point in the FP tree and the value of the node in the second category.

[0043] The processes corresponding to the nodes in the second category can be defaulted to the processes that some abnormal kidney disease biological samples have gone through. If the kidney disease biological samples are abnormal due to the processes corresponding to the nodes in the second category, these processes should also be obtained through the FP tree; at this time, although the processes corresponding to the nodes in the second category are not the processes that all abnormal kidney disease biological samples have gone through, they also reflect the more obvious abnormal characteristics of some abnormal kidney disease biological samples.

[0044] In order to analyze the possibility that the process links of the partial path corresponding to the second category cause abnormal kidney disease biological samples and avoid missing the process links that cause abnormal kidney disease biological samples, the node with the smallest value in the second category is first taken as the end point of the target path. When the value of the end point has a high similarity with the values ​​of other nodes in the second category, it means that the process corresponding to the end point and the processes corresponding to other nodes in the second category always appear together. The stronger the correlation between the processes of the partial path corresponding to the second category, the greater the possibility that the process links of the partial path corresponding to the second category jointly cause abnormal kidney disease biological samples.

[0045] On the other hand, if the value of the next node at the end point has a high similarity with the value of the node in the second category, it means that the process corresponding to the next node not included in the target path also has a high correlation with the process of the target path, which indirectly indicates that the judgment of the associated process obtained by the maximum frequent item set corresponding to the target path is incomplete, which further reflects that the process link of the path corresponding to the second category is less likely to be the process link that causes abnormal kidney disease biological samples.

[0046] Therefore, this embodiment obtains the second abnormality possibility degree of the second category according to the fluctuation of the value of the node in the second category and the similarity between the value of the node at the next depth of the end point in the FP tree and the value of the node in the second category. The greater the second abnormality possibility degree, the greater the possibility that the process link of the path corresponding to the second category is the process link that causes the abnormality of the kidney disease biological sample.

[0047] Preferably, in a method that can be implemented in this embodiment, the method for obtaining the second abnormal possibility degree is: the standard deviation of the values ​​of the nodes in the second category is negatively correlated and normalized as the result, as the initial abnormal possibility degree of the second category; the greater the initial abnormal possibility degree, the more equal the values ​​of the nodes in the second category, which indirectly indicates that the process of the corresponding path of the second category jointly causes the possibility of abnormality in the kidney disease biological sample. In order to analyze the integrity of the maximum frequent item set corresponding to the target path, the variance of the values ​​of the nodes in the second category is obtained as the first variance; the variance of the values ​​of all nodes at the next depth of the depth where the end point is located in the FP tree and the values ​​of the nodes in the second category is obtained as the second variance; when the first variance is more equal to the second variance, it means that the maximum frequent item set corresponding to the target path is more incomplete, which indirectly indicates that the path corresponding to the second category is more incomplete, and the process links of the path corresponding to the second category jointly cause the possibility of abnormality in the kidney disease biological sample. Furthermore, in this embodiment, the result of normalizing the absolute value of the difference between the first variance and the second variance is used as the correction weight of the second category. The larger the correction weight, the more complete the path corresponding to the second category, and the greater the possibility that the process links of the path corresponding to the second category jointly cause abnormalities in the kidney disease biological sample; and the product of the correction weight of the second category and the initial abnormal possibility degree is used as the second abnormal possibility degree of the second category.

[0048] The calculation formula for the second abnormal possibility degree is: ; Where p is the second abnormality possibility of the second category; is the first variance; is the second variance; is the absolute value function; is the modified weight for the second category; is the standard deviation of the values ​​of the nodes in the second category; is the initial abnormal possibility of the second category; norm is the normalization function; exp is the exponential function with a natural constant as the base.

[0049] Step S205: acquiring the overall abnormality degree of the maximum frequent itemset according to the similarity degree, the first abnormality possibility degree and the second abnormality possibility degree.

[0050] It is known that the greater the similarity, the greater the possibility that the process link corresponding to the target path causes abnormal kidney disease biological samples; the greater the first abnormal possibility, the greater the possibility that the process link of the path corresponding to the first category is that abnormal kidney disease biological samples frequently pass through the process link together, which indirectly indicates that the process link corresponding to the target path causes abnormal kidney disease biological samples. The greater the second abnormal possibility, the greater the possibility that the process link of the path corresponding to the second category causes abnormal kidney disease biological samples together, which also indirectly indicates that the process link corresponding to the target path causes abnormal kidney disease biological samples. The greater the overall abnormality, the more likely the process in the maximum frequent item set is to be the process that causes abnormal kidney disease biological samples.

[0051] Preferably, in a manner that can be implemented in this embodiment, the method for obtaining the overall abnormality degree is: obtaining the average of the first abnormality possibility degree and the second abnormality possibility degree as the overall abnormality possibility degree of the maximum frequent item set; and taking the product of the similarity degree and the overall abnormality possibility degree as the overall abnormality degree of the maximum frequent item set.

[0052] The calculation formula for the overall abnormality degree is: ; In the formula, r is the overall abnormality degree of the maximum frequent item set; S is the similarity degree of the nodes on the target path; q is the first abnormality possibility degree of the first category; p is the second abnormality possibility degree of the second category; is the overall abnormal possibility of the maximum frequent itemset.

[0053] Step S206: Obtain the importance of the maximum frequent itemset according to the depth of the end point in the FP tree and the value of each node in the target path.

[0054] When the depth of the target path is greater, it means that there are some processes with less correlation in the maximum frequent item set, and the processes in the maximum frequent item set cannot represent the processes that abnormal kidney disease biological samples have passed through together, and the importance of the maximum frequent item set is smaller; when the node value on the target path of the maximum frequent item set is larger, it means that the processes in the maximum frequent item set are more likely to be the processes that abnormal kidney disease biological samples have passed through together, and the possibility that the processes in the maximum frequent item set cause abnormal kidney disease biological samples is greater, and the importance of the maximum frequent item set is greater. Therefore, this embodiment obtains the importance of the maximum frequent item set according to the depth of the end point in the FP tree and the value of each node in the target path.

[0055] Preferably, in an implementable manner of this embodiment, the importance is obtained by taking the ratio of the mean value of all nodes in the target path to the depth of the end point in the FP tree as the importance of the maximum frequent itemset.

[0056] Step S207: taking the product of the overall abnormality degree and the importance degree as the reference weight of the maximum frequent itemset.

[0057] If there are fewer processes in the maximum frequent itemset, that is, the target path corresponding to the maximum frequent itemset is shorter, after the target path is classified, the number of nodes in each category is relatively small, so that the calculated overall abnormality degree of the maximum frequent itemset cannot truly reflect the abnormal process corresponding to the abnormal kidney disease biological sample. Therefore, this embodiment combines the importance of the maximum frequent itemset, that is, obtains the product of the overall abnormality degree of the maximum frequent itemset and the importance, as the reference weight of the maximum frequent itemset.

[0058] At this point, the reference weight of each maximum frequent itemset is obtained.

[0059] Step S3: According to the reference weight, obtain the rationality of the support threshold.

[0060] In order to comprehensively analyze the possibility of abnormal kidney disease biological samples caused by the process link corresponding to the maximum frequent item set obtained by the support threshold, the mean of the reference weights of all maximum frequent item sets is used as the rationality of the support threshold. The greater the rationality, the more reasonable the support threshold, and the process link corresponding to the maximum frequent item set corresponding to the support threshold is more likely to cause abnormal kidney disease biological samples.

[0061] At this point, the reasonableness of each support threshold is obtained.

[0062] Step S4: obtaining an optimal support threshold based on the reasonableness, and predicting abnormal process links of kidney disease biological samples according to the optimal support threshold.

[0063] Specifically, in order to avoid missing the process link that causes abnormal kidney disease biological samples, this embodiment uses the support threshold corresponding to the maximum reasonable degree as the optimal support threshold. The FP tree constructed by the process of abnormal kidney disease biological sample circulation is processed by the optimal support threshold, and each maximum frequent item set obtained at this time is used as the target item set; wherein the process link of the path corresponding to each target item set is the process link that may cause abnormal kidney disease biological samples.

[0064] In order to accurately determine the abnormal process link that causes the abnormality of the kidney disease biological sample, this embodiment further obtains the flow process of each normal kidney disease biological sample in the batch of kidney disease biological samples and constructs an FP tree as a reference FP tree, processes the reference FP tree through the optimal support threshold, and uses each maximum frequent item set obtained at this time as a reference item set; then compares each target item set with each reference item set, and uses the process link of the path corresponding to the target item set that is not the same as the reference item set as the abnormal process link of the kidney disease biological sample.

[0065] At this point, the abnormal process links that are likely to cause abnormalities in kidney disease biological samples have been accurately predicted. Before the kidney disease biological samples pass through the predicted abnormal process links, the temperature, humidity and other environments of the kidney disease biological samples can be strictly controlled, and the behaviors that cause abnormalities in the kidney disease biological samples can be avoided. The abnormal process links can also be processed and risks avoided in advance, effectively reducing abnormalities in the kidney disease biological samples during transportation.

[0066] In summary, this embodiment obtains the flow process of abnormal kidney disease biological samples and constructs an FP tree; for any support threshold in the preset support threshold set, the FP tree is processed by the support threshold to obtain the maximum frequent item set; based on the value and depth of each node in the FP tree corresponding to the path of the maximum frequent item set, the reference weight of the maximum frequent item set is obtained, and then the rationality of the support threshold is obtained; based on the rationality, the optimal support threshold is obtained, and the abnormal process link of the kidney disease biological sample is predicted according to the optimal support threshold. The present invention obtains the optimal support threshold, and then accurately predicts the flow process that is likely to cause abnormal kidney disease biological samples, and promptly handles and avoids abnormal risks, effectively reducing the abnormal phenomenon of kidney disease biological samples.

[0067] Embodiment 2: The present invention also proposes a sample information traceability system for a kidney disease biological sample library, please refer to Figure 3, which shows a structure diagram of a sample information traceability system of a kidney disease biological sample library provided by an embodiment of the present invention, the system includes: a maximum frequent item set acquisition module 10, a reference weight acquisition module 20, a rationality acquisition module 30 and an abnormal process link acquisition module 40.

[0068] The maximum frequent item set acquisition module 10 is used to acquire the flow process of each abnormal kidney disease biological sample and construct an FP tree; for any support threshold in the preset support threshold set, the FP tree is processed by the support threshold to acquire the maximum frequent item set.

[0069] The reference weight acquisition module 20 is used to acquire the reference weight of each maximum frequent item set based on the value and depth of each node of the path corresponding to each maximum frequent item set in the FP tree.

[0070] The reasonableness acquisition module 30 is used to acquire the reasonableness of the support threshold according to the reference weight.

[0071] The abnormal process link acquisition module 40 is used to obtain the optimal support threshold based on the rationality, and predict the abnormal process link of the kidney disease biological sample according to the optimal support threshold.

[0072] It should be noted that: the system provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device is divided into different functional modules to complete all or part of the functions described above. In addition, the sample information traceability system of a kidney disease biological sample library provided in the above embodiment and the sample information traceability method embodiment of a kidney disease biological sample library belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0073] Embodiment 3: The present invention also proposes a sample information tracing device for a kidney disease biological sample library, the device comprising a memory and a processor, wherein the memory stores an executable program code, and the processor is used to call and execute the executable program code to execute a sample information tracing method for a kidney disease biological sample library provided in an embodiment of the present application. The device can be specifically a chip, a component or a module, and the chip may include a connected processor and a memory; wherein the memory is used to store instructions, and when the processor calls and executes the instructions, the chip can execute a sample information tracing method for a kidney disease biological sample library provided in the above embodiment.

[0074] In addition, the present application embodiment also protects a computer device, see Figure 4The computer device includes a memory 401, a processor 402, and a computer program 403 stored in the memory 401 and running on the processor 402, wherein when the processor 402 executes the computer program 403, the computer device can execute any of the sample information tracing methods of the kidney disease biological sample library introduced above.

[0075] Embodiment 4: This embodiment also provides a computer-readable storage medium, in which a computer program code is stored. When the computer program code is executed on a computer, the computer executes the above-mentioned related method steps to implement a sample information tracing method for a kidney disease biological sample library provided in the above embodiment.

[0076] Embodiment 5: This embodiment also provides a computer program product. When the computer program product is run on a computer, the computer is caused to execute the above-mentioned related steps to implement a method for tracing sample information of a kidney disease biological sample library provided in the above embodiment.

[0077] Among them, the device, computer-readable storage medium, computer program product or chip provided in this embodiment is used to execute the corresponding method provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method provided above, and will not be repeated here.

[0078] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0079] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for tracing sample information of a kidney disease biological sample library, characterized in that: The method comprises the following steps: Obtain the flow process of each abnormal kidney disease biological sample and construct an FP tree; for any support threshold in the preset support threshold set, process the FP tree through the support threshold to obtain the maximum frequent item set; Based on the value and depth of each node in the FP tree corresponding to each maximum frequent item set, obtain the reference weight of each maximum frequent item set; According to the reference weight, obtaining the rationality of the support threshold; The optimal support threshold is obtained based on the reasonable degree, and the abnormal process links of kidney disease biological samples are predicted according to the optimal support threshold.

2. The method for tracing the sample information of a kidney disease biological sample library according to claim 1, characterized in that: The method for obtaining the reference weight is: For any maximum frequent item set, the path corresponding to the maximum frequent item set is used as the target path, and the nodes in the target path are divided according to the value size to obtain the first category and the second category; wherein the value of the node in the first category is larger, and the value of the node in the second category is smaller; According to the numerical difference between the nodes in the first category and the nodes in the second category, the similarity of the nodes on the target path is obtained; The node with the largest value in the second category is used as the split node of the target path, and the first abnormality possibility degree of the first category is obtained according to the numerical difference between the split node and other nodes at the same depth in the FP tree, and the number of all nodes at the depth of the split node in the FP tree; The node with the smallest value in the second category is taken as the end point of the target path, and the second abnormal possibility degree of the second category is obtained according to the fluctuation of the value of the node in the second category and the similarity between the value of the node at the next depth of the end point in the FP tree and the value of the node in the second category; According to the similarity degree, the first abnormal possibility degree and the second abnormal possibility degree, obtaining the overall abnormality degree of the maximum frequent item set; Obtaining the importance of the maximum frequent itemset according to the depth of the end point in the FP tree and the value of each node in the target path; The product of the overall abnormality degree and the importance degree is used as the reference weight of the maximum frequent item set.

3. The method for tracing the sample information of a kidney disease biological sample library according to claim 2, characterized in that: The method for obtaining the similarity degree is: Obtaining the average value of the values ​​of all nodes in the first category as a first reference value; Obtaining the average value of the values ​​of all nodes in the second category as a second reference value; The result of negatively correlating and normalizing the difference between the first reference value and the second reference value is used as the similarity of the nodes on the target path.

4. The method for tracing the sample information of a kidney disease biological sample library according to claim 2, characterized in that: The method for obtaining the first abnormality possibility degree is: The depth of the split node in the FP tree is taken as the target depth, and the average value of the values ​​of all nodes except the split node at the target depth is obtained as the first eigenvalue; The difference between the value of the segmentation node and the first characteristic value is used as a reference abnormality degree of the first category; The result of normalizing the product of the inverse of the number of all nodes at the target depth and the reference abnormality degree is taken as the first abnormality possibility degree of the first category.

5. The method for tracing the sample information of a kidney disease biological sample library according to claim 2, characterized in that: The method for obtaining the second abnormal possibility degree is: The result of negatively correlating and normalizing the standard deviation of the values ​​of the nodes in the second category is used as the initial abnormal possibility degree of the second category; Obtain the variance of the values ​​of the nodes in the second category as the first variance; Obtaining the variance of the values ​​of all nodes at the next depth of the end point in the FP tree and the values ​​of the nodes in the second category as the second variance; The result of normalizing the absolute value of the difference between the first variance and the second variance is used as the correction weight of the second category; The product of the revised weight of the second category and the initial abnormality possibility degree is taken as the second abnormality possibility degree of the second category.

6. The method for tracing the sample information of a kidney disease biological sample library according to claim 2, characterized in that: The method for obtaining the overall abnormality degree is: Obtaining the average of the first abnormality possibility degree and the second abnormality possibility degree as the overall abnormality possibility degree of the maximum frequent item set; The product of the similarity degree and the overall abnormality possibility degree is taken as the overall abnormality degree of the maximum frequent item set.

7. The method for tracing the sample information of a kidney disease biological sample library according to claim 2, characterized in that: The method for obtaining the importance is: The ratio of the average value of all nodes in the target path to the depth of the end point in the FP tree is used as the importance of the maximum frequent itemset.

8. The method for tracing the sample information of a kidney disease biological sample library according to claim 1, characterized in that: The method for obtaining the reasonable degree is: The mean of the reference weights of all maximum frequent itemsets is taken as the reasonableness of the support threshold.

9. The method for tracing the sample information of a kidney disease biological sample library according to claim 1, characterized in that: The method for obtaining the optimal support threshold is: The support threshold corresponding to the maximum reasonable degree is taken as the optimal support threshold.

10. The method for tracing the sample information of a kidney disease biological sample library according to claim 1, characterized in that: The method for obtaining the abnormal process link is: The FP tree is processed by the optimal support threshold, and each maximum frequent item set obtained at this time is used as the target item set; Obtain the flow process of each normal kidney disease biological sample and construct an FP tree as a reference FP tree, process the reference FP tree through the optimal support threshold, and use each maximum frequent item set obtained at this time as a reference item set; Each target item set is compared with each reference item set, and the process links of the paths corresponding to the target item sets that are different from the reference item sets are regarded as abnormal process links of the kidney disease biological samples.

Images