Drug non-clinical research test report automatic generation system

Abstract

The invention discloses an automatic generation system for a non-clinical drug research test report, and relates to the technical field of test data processing. Comprising a data extraction module used for extracting original data, a preprocessing module used for performing data preprocessing according to test numbers, a statistical analysis module used for reading the preprocessed data and performing statistical analysis, and a data warehouse used for storing the preprocessed data and statistical analysis results, the system comprises a visual chart generation module used for generating a visual chart, an abnormity analysis module used for carrying out abnormal value extraction and judgment analysis, a text report generation module used for compiling a text report, and a customer service system module used for displaying the corresponding visual chart and text report on a display interface. According to the invention, full automation of the non-clinical research test data processing and analysis process of drugs can be realized, and visual and readable visual charts and text reports can be displayed through a customer service system interface.

Description

Technical Field

[0001] The present invention relates to the technical field of test data processing, and in particular to a system for automatically generating drug non-clinical research test reports. Background Art

[0002] Laboratories conducting various types of non-clinical research trials for drugs may use multiple animal test management systems from different manufacturers, some of which are foreign systems such as Pristima and Waston systems, which involve a large amount of statistical processing of test data. The test data is scattered on different systems, and the data storage formats are different. It is necessary to use computer technology to merge the data for unified processing and presentation. The existing management systems are all closed systems, and if customers want to view the test data generated during the test, the previous model is to export the report in the management system and then email it to the customer, which is not conducive to the customer's timely control of the test status.

[0003] The current statistical work method is: (1) manually export data formats such as PDF and CSV from different management systems; (2) manually import into statistical software such as STAT, write statistical methods, and obtain statistical results; (3) manually make an appendix file; (4) perform quality inspection on the above process.

[0004] The existing problems are: exporting data from different management systems easily causes management system jams, requires manual participation, heavy workload, and the overall process is time-consuming. For example, to complete the preparation of a regular long-term toxicity test statistical appendix: steps 1 to 3 require a skilled person's one-day work time, and the quality inspection of step 4 takes nearly one day. It takes a total of 2 days. Summary of the invention

[0005] Based on the above problems, the purpose of the present invention is to provide an automatic generation system for non-clinical research test reports of drugs, to realize full automation of test data processing and analysis process, and to display intuitive and easy-to-read visual charts and text reports through the customer service system interface, so as to meet the needs of end users to remotely and timely control the test status.

[0006] The technical solution adopted by the present invention to achieve its invention object is a system for automatically generating drug non-clinical research test reports, comprising:

[0007] Data extraction module, used to extract raw data from various animal test management systems;

[0008] A preprocessing module, used to preprocess the raw data extracted by the data extraction module according to the test number, and load the preprocessed data into a data warehouse;

[0009] A statistical analysis module is used to read the pre-processed data in the data warehouse and perform statistical analysis, and write the statistical analysis results into the data warehouse, wherein the statistical analysis includes calculating the mean change rate of each dose group compared with the control group, and performing statistical analysis on the difference of quantitative indicators;

[0010] A data warehouse, used to store the preprocessed data and the statistical analysis results;

[0011] A chart generation module, for extracting the statistical analysis results from the data warehouse and generating selected indicators of the selected test into visual charts;

[0012] An abnormality analysis module, used to extract the statistical analysis results from the data warehouse and perform abnormal value extraction and judgment analysis;

[0013] A text report generation module, used to compile a text report based on the results output by the abnormality analysis module;

[0014] The customer service system module is used to retrieve the visual chart and text report, query and display the corresponding visual chart and text report on the display interface according to the test number, and export electronic files. Multiple customer service system modules can be remotely set up at the end user's location according to user needs.

[0015] Furthermore, the raw data includes formatted data and unformatted data, and the data preprocessing includes removing duplicate and erroneous data, which have been marked in the animal experiment management system, and filling in missing experiment numbers for unformatted data to ensure the data quality and accuracy of the data warehouse.

[0016] Furthermore, the pre-processed data stored in the data warehouse includes multiple administration test data, reproductive toxicity test data, and carcinogenicity test data, wherein:

[0017] The data of multiple-dose study included body weight, food intake, clinical examination, organs, general observations, and gross anatomical observations;

[0018] Reproductive toxicity test data include body weight, food intake, organs, general observations and gross anatomical observations, sexual cycle, mating records, pregnant animal weight, food intake, embryo formation and fetal physiological indicators;

[0019] Carcinogenicity test data included body weight, food intake, organs, general and gross anatomical observations, presence of neoplastic lesions, survival-mortality, and pathology.

[0020] Furthermore, the statistical analysis module calculates the mean change rate of each dosage group compared with the control group as follows:

[0021]

[0022] Furthermore, the statistical analysis module performs statistical analysis on the quantitative indicators in the following manner:

[0023] First, the sample data is tested for prerequisites, and the test samples are randomly grouped to meet independence;

[0024] The Shapiro-Wilk test was used to test whether the data set obeyed the normal distribution, and the LEVENE test was used for the variance homogeneity test. If the LEVENE test showed P>0.05, it meant that there was no statistical difference, and ANOVA analysis of variance was used for statistical analysis; if the LEVENE test showed P≤0.05, it meant that the difference was statistically significant, and the Kruskal-Wallis H rank sum test was used for statistical difference analysis;

[0025] When the ANOVA analysis of variance showed P ≤ 0.05, it indicated that the difference was statistically significant, and a specific analysis of the difference was performed, using the Dunnett method for multiple comparisons to compare the differences between the groups; when the ANOVA analysis of variance showed P > 0.05, it indicated that there was no statistical difference, and the statistics were terminated;

[0026] When the Kruskal-Wallis H rank sum test showed P ≤ 0.05, it indicated that the difference was statistically significant, and the Mann-Whitney U test was used to compare the differences between the groups; when the Kruskal-Wallis H rank sum test showed P> 0.05, it indicated that there was no statistical difference, and the statistics were terminated;

[0027] The comparison of inter-group differences was conducted between each dose group and the control group. All tests were two-sided at 0.05, that is, the significance level of the two-sided test was set at 0.05. It refers to a statistical test method used when the purpose of statistical analysis is to test whether there is a significant difference between the sample mean and the overall mean, or the sample proportion, but does not ask whether the direction of the difference is positive or negative.

[0028] For example, the null hypothesis: Group 1 = Group 2, or the alternative hypothesis: Group 1 ≠ Group 2 is two-sided, and if Group 1 < Group 2, it is one-sided.

[0029] The test is based on sample data to determine whether to reject the null hypothesis that Group 1 and Group 2 are the same. After performing the hypothesis test, the judgment is made based on the obtained p-value. If the p-value is less than the significance level of 0.05, the null hypothesis can be rejected and there is a statistically significant difference between Group 1 and Group 2.

[0030] The detection indicators such as body weight and food intake that are numerical values ​​are quantitative indicators.

[0031] Furthermore, the statistical analysis result data table output by the statistical analysis module written into the data warehouse contains the following fields:

[0032] Test number STUDY_NUMBER, group GROUPID, sex SEX, mean MEAN, standard deviation DEVIATION, P value, number of animals ANIMALNUM, indicator parameter PARA_NAME, unit UNIT, test period TP, date RECORDDATE, mean change rate RATIOS and type TYPENAME.

[0033] Furthermore, the statistical analysis module includes multiple statistical analysis submodules, which are respectively used for parallel statistical analysis of body weight, food intake, clinical examination, organs, general observation, gross anatomical observation, sexual cycle, mating record, pregnant animal weight, food intake, embryo formation and fetal physiological indicators, the presence of tumor lesions, survival-mortality and pathological test data. Multiple statistical analysis submodules are processed in parallel, which can greatly improve the system processing speed.

[0034] Furthermore, the abnormal analysis module performs abnormal value extraction and judgment analysis specifically as follows: descriptive statistics are performed on qualitative indicators, and abnormal symptoms are listed; abnormal values ​​are found and marked for quantitative indicators, and the abnormal values ​​include the mean change rate of each dose group compared with the control group is greater than 10% or less than -10%, when the statistical difference test P value is less than or equal to 0.05, there is a dose correlation, and the mean value exceeds the background data confidence interval.

[0035] The detection indicators are observational, to a certain extent or descriptive, such as general observation, which are qualitative indicators.

[0036] The data extraction module, preprocessing module, statistical analysis module, chart generation module, abnormal analysis module, and text report generation module can all be set to run automatically at a scheduled time.

[0037] The beneficial effects of the present invention are:

[0038] The system can automatically and regularly run data extraction, statistical analysis, chart and text report generation operations to ensure the accuracy and efficiency of test data collation and statistical analysis. It solidifies a series of manual work processes into automatic programs, and automatically extracts raw data from existing animal test management systems for processing every day. The extraction time can be set freely, and can be set to idle time, such as starting the extraction at night, which will not cause the existing animal test management system to freeze. It realizes the automation of research test data processing and analysis process, greatly reduces manual workload, saves a lot of processing time, and can display intuitive and easy-to-read visual charts and text reports through the customer service system interface, which can meet the end user's needs for remote and timely control of the test status. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1This is a schematic diagram of the system framework of an embodiment of the present invention;

[0040] Figure 2 A schematic diagram of a process of performing difference statistical analysis by a statistical analysis module according to an embodiment of the present invention;

[0041] Figure 3 A schematic diagram of a visualization chart displayed on the customer service system interface of an embodiment of the present invention;

[0042] Figure 4 This is a schematic diagram of a text report displayed on the customer service system interface of an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0044] Figures 1 to 4 A specific embodiment of the system for automatically generating a drug non-clinical research test report of the present invention is shown, comprising:

[0045] A data extraction module, using an extraction, transformation and loading (ETL) tool in the prior art, is used to extract the original data of the Pristma animal test management system; the original data includes formatted data and unformatted data;

[0046] A preprocessing module is used to preprocess the raw data extracted by the data extraction module according to the test number, and load the preprocessed data into the data warehouse; the data preprocessing includes removing duplicate and erroneous data, which have been marked in the Pristma animal test management system, and filling in the missing test number for the non-formatted data;

[0047] The pre-treatment data include multiple administration test data, reproductive toxicity test data, and carcinogenicity test data, wherein:

[0048] The data of multiple-dose study included body weight, food intake, clinical examination, organs, general observations, and gross anatomical observations;

[0049] Reproductive toxicity test data include body weight, food intake, organs, general observations and gross anatomical observations, sexual cycle, mating records, pregnant animal weight, food intake, embryo formation and fetal physiological indicators;

[0050] Carcinogenicity test data included body weight, food intake, organs, general and gross anatomical observations, presence of neoplastic lesions, survival-mortality, and pathology.

[0051] Among them, body weight, food intake, organs, general observation and gross anatomical observation are test index data required for multiple dose administration tests, reproductive toxicity tests and carcinogenicity tests.

[0052] Statistical analysis module, this embodiment includes thirteen statistical analysis submodules, which are respectively used for parallel statistical analysis of thirteen kinds of experimental data, including body weight, food intake, clinical examination, organs, general observation, gross anatomical observation, sexual cycle, mating record, pregnant animal weight, food intake, embryo formation and fetal physiological indicators, the presence of tumor lesions, survival-mortality and pathology. Use R language to write statistical analysis submodules according to each indicator, which are used to read the pre-processed data in the data warehouse and perform statistical analysis, and write the statistical analysis results into the data warehouse; the statistical analysis includes calculating the mean change rate of each dose group compared with the control group, and performing differential statistical analysis on quantitative indicators; the statistical analysis result data table contains the fields: test number STUDY_NUMBER, group GROUPID, sex SEX, mean MEAN, standard deviation DEVIATION, P value, animal number ANIMALNUM, indicator parameter PARA_NAME, unit UNIT, test period TP, date RECORDDATE, mean change rate RATIOS and type TYPENAME;

[0053] A data warehouse, used to store the preprocessed data and the statistical analysis results;

[0054] The chart generation module uses existing business intelligence BI tools, such as Fanruan BI, to extract the statistical analysis results from the data warehouse and generate the selected indicators of the selected test into visual charts; the TYPENAME field in the statistical analysis result data table can distinguish which indicator is which test, and perform chart template design, display interface design, custom selection functions, etc. according to each test and each indicator; for example, multiple drug administration weight data chart template: weight is a quantitative indicator, and the mean ± standard deviation (X ± SD) of each group and the mean of each dose group compared with the control group are displayed in the table Change rate (%), and weight is continuous data that changes over time, so a line graph is used to show trend changes. In the line graph, time points are evenly distributed along the horizontal axis, and weight values ​​are evenly distributed along the vertical axis. Various lines represent different groups. For example, if general observation records are qualitative indicators, the count and incidence of the observed symptoms in each group are displayed in the table. In order to show the proportion of abnormal symptoms in each group, a grouped stacked bar chart is used. The custom selection function can display the corresponding result chart based on the manually selected and filled parameters such as test number, indicator, test cycle, dose group, etc.

[0055] The abnormal analysis module is used to extract the statistical analysis results from the data warehouse and perform abnormal value extraction and judgment analysis, specifically: perform descriptive statistics on qualitative indicators and list abnormal symptoms; find abnormal values ​​for quantitative indicators and mark them, the abnormal values ​​include the mean change rate of each dose group compared with the control group is greater than 10% or less than -10%, when the statistical difference test P value is less than or equal to 0.05, there is a dose correlation, and the mean value exceeds the background data confidence interval; the background data is obtained by sorting and statistically analyzing millions of various test indicator data generated by the tests conducted in the past five years, and is used to determine the reliable reference interval of the laboratory test quantitative indicators, that is, the confidence interval, which is fully useful for comparing whether the test indicators are abnormal;

[0056] A text report generation module, using existing business intelligence BI tools, such as Fanruan BI, to compile a text report based on the results output by the anomaly analysis module;

[0057] The customer service system module is used to retrieve the visual charts and text reports, query according to the test number and display the corresponding visual charts and text reports on the display interface, and export electronic files in various formats such as HTML, XLSX, MS Offices, PDF, etc. for users to download.

[0058] In this embodiment, the statistical analysis module calculates the mean change rate of each dosage group compared with the control group as follows:

[0059]

[0060] The statistical analysis module performs statistical analysis on the quantitative indicators in the following ways:

[0061] First, the sample data is tested for prerequisites, and the test samples are randomly grouped to meet independence;

[0062] The Shapiro-Wilk test was used to test whether the data set obeyed the normal distribution, and the LEVENE test was used for the variance homogeneity test. If the LEVENE test showed P>0.05, it meant that there was no statistical difference, and ANOVA analysis of variance was used for statistical analysis; if the LEVENE test showed P≤0.05, it meant that the difference was statistically significant, and the Kruskal-Wallis H rank sum test was used for statistical difference analysis;

[0063] When the ANOVA analysis of variance showed P ≤ 0.05, it indicated that the difference was statistically significant, and a specific analysis of the difference was performed, using the Dunnett method for multiple comparisons to compare the differences between the groups; when the ANOVA analysis of variance showed P > 0.05, it indicated that there was no statistical difference, and the statistics were terminated;

[0064] When the Kruskal-Wallis H rank sum test showed P ≤ 0.05, it indicated that the difference was statistically significant, and the Mann-Whitney U test was used to compare the differences between the groups; when the Kruskal-Wallis H rank sum test showed P> 0.05, it indicated that there was no statistical difference, and the statistics were terminated;

[0065] The inter-group differences were compared between each dose group and the control group, and all tests were two-sided at 0.05.

[0066] In the present invention, the data extraction module, preprocessing module, statistical analysis module, data warehouse, chart generation module, abnormal analysis module, and text report generation module are stored and run on the laboratory server as program modules or database modules, and the customer service system module is stored and run on the client computer as a program module.

[0067] By using the embodiment of the present invention to perform automatic statistical analysis, nearly 1 million pieces of raw weight test data from all laboratory tests were screened and statistical results were generated. Within 15 minutes, 50,000 pieces of result data were generated after sorting and statistics were completed, realizing automatic and rapid extraction of test data, data statistical analysis, data visualization, and report automation.

[0068] The above embodiments of the present invention are merely examples for illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes and modifications can be made based on the above description. It is impossible to list all the embodiments here. Any obvious changes or modifications derived from the technical solution of the present invention are still within the scope of protection of the present invention.

Claims

1. A system for automatically generating drug non-clinical research test reports, characterized in that: include: Data extraction module, used to extract raw data from various animal test management systems; A preprocessing module, used to preprocess the raw data extracted by the data extraction module according to the test number, and load the preprocessed data into a data warehouse; A statistical analysis module is used to read the pre-processed data in the data warehouse and perform statistical analysis, and write the statistical analysis results into the data warehouse, wherein the statistical analysis includes calculating the mean change rate of each dose group compared with the control group, and performing statistical analysis on the difference of quantitative indicators; A data warehouse, used to store the preprocessed data and the statistical analysis results; A chart generation module, for extracting the statistical analysis results from the data warehouse and generating selected indicators of the selected test into visual charts; An abnormality analysis module, used to extract the statistical analysis results from the data warehouse and perform abnormal value extraction and judgment analysis; A text report generation module, used to compile a text report based on the results output by the abnormality analysis module; The customer service system module is used to retrieve the visual charts and text reports, query according to the test number and display the corresponding visual charts and text reports on the display interface, and export electronic files.

2. The system for automatically generating drug non-clinical research test reports according to claim 1, characterized in that: The raw data includes formatted data and unformatted data, and the data preprocessing includes removing duplicate and erroneous data and filling in missing test numbers for unformatted data.

3. The system for automatically generating drug non-clinical research test reports according to claim 1, characterized in that: The pre-processed data stored in the data warehouse include multiple administration test data, reproductive toxicity test data, and carcinogenicity test data, wherein: The data of multiple-dose study included body weight, food intake, clinical examination, organs, general observations, and gross anatomical observations; Reproductive toxicity test data include body weight, food intake, organs, general observations and gross anatomical observations, sexual cycle, mating records, pregnant animal weight, food intake, embryo formation and fetal physiological indicators; Carcinogenicity test data included body weight, food intake, organs, general and gross anatomical observations, presence of neoplastic lesions, survival-mortality, and pathology.

4. The system for automatically generating drug non-clinical research test reports according to claim 3, characterized in that: The statistical analysis module calculates the mean change rate of each dosage group compared with the control group as follows:

5. The system for automatically generating drug non-clinical research test reports according to claim 3, characterized in that: The statistical analysis module performs statistical analysis on the quantitative indicators in the following ways: First, the sample data is tested for prerequisites, and the test samples are randomly grouped to meet independence; The Shapiro-Wilk test was used to test whether the data set obeyed the normal distribution, and the LEVENE test was used for the variance homogeneity test. If the LEVENE test showed P>0.05, it meant that there was no statistical difference, and ANOVA analysis of variance was used for statistical analysis; if the LEVENE test showed P≤0.05, it meant that the difference was statistically significant, and the Kruskal-Wallis H rank sum test was used for statistical difference analysis; When the ANOVA analysis of variance showed P ≤ 0.05, it indicated that the difference was statistically significant, and a specific analysis of the difference was performed, using the Dunnett method for multiple comparisons to compare the differences between the groups; when the ANOVA analysis of variance showed P > 0.05, it indicated that there was no statistical difference, and the statistics were terminated; When the Kruskal-Wallis H rank sum test showed P ≤ 0.05, it indicated that the difference was statistically significant, and the Mann-Whitney U test was used to compare the differences between the groups; when the Kruskal-Wallis H rank sum test showed P> 0.05, it indicated that there was no statistical difference, and the statistics were terminated; The inter-group differences were compared between each dose group and the control group, and all tests were two-sided at 0.

05.

6. The system for automatically generating drug non-clinical research test reports according to claim 3, characterized in that: The statistical analysis result data table output by the statistical analysis module written into the data warehouse contains the following fields: Test number STUDY_NUMBER, group GROUPID, sex SEX, mean MEAN, standard deviation DEVIATION, P value, number of animals ANIMALNUM, indicator parameter PARA_NAME, unit UNIT, test period TP, date RECORDDATE, mean change rate RATIOS and type TYPENAME.

7. The system for automatically generating drug non-clinical research test reports according to claim 3, characterized in that: The statistical analysis module includes multiple statistical analysis submodules, which are used for parallel statistical analysis of body weight, food intake, clinical examination, organs, general observation, gross anatomical observation, sexual cycle, mating record, pregnant animal weight, food intake, embryo formation and fetal physiological indicators, the presence of tumor lesions, survival-mortality and pathological test data.

8. The system for automatically generating drug non-clinical research test reports according to claim 3, characterized in that: The abnormal analysis module performs abnormal value extraction and judgment analysis specifically as follows: descriptive statistics are performed on qualitative indicators, and abnormal symptoms are listed; abnormal values ​​are found and marked for quantitative indicators, and the abnormal values ​​include the mean change rate of each dose group compared with the control group is greater than 10% or less than -10%, when the statistical difference test P value is less than or equal to 0.05, there is a dose correlation, and the mean value exceeds the background data confidence interval.

Images