Screening, application and storage processing of public place environment key microorganisms based on machine learning
By using machine learning technology, combined with high-throughput genome sequencing and random forest models, the problem of identifying pathogenic microorganisms in public places has been solved, enabling the screening and monitoring of key pathogenic microorganisms and improving data analysis efficiency and environmental safety.
Patent Information
- Application Number
- CN202411029258.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-30
- Publication Date
- 2024-11-26
AI Technical Summary
Existing technologies have not yet been able to effectively utilize machine learning to identify and screen key pathogenic microorganisms in public places, especially in the massive sequencing data, where it is difficult to find key pathogenic microorganisms and cannot meet the seasonal characteristics monitoring needs of pathogenic microorganisms in public places.
Using a machine learning-based approach, a key microbial screening and processing workflow was established through high-throughput genome sequencing, random forest model, and 10-fold cross-validation. This workflow included sample collection, data cleaning, model training, and feature selection to identify key microbial indicators.
It improves the efficiency and utilization value of microbial data in public places, identifies key pathogenic microorganisms, provides characteristic monitoring targets for the monitoring and early warning of pathogenic microorganisms in the environment, and enhances the safety of public places.
Smart Images

Figure CN119028432A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of public place environment detection, and in particular to the field of microbial detection, and specifically refers to a method, device, processor and computer-readable storage medium thereof for realizing key microbial identification, screening and processing in public place environments based on machine learning. Background Art
[0002] As an important infrastructure of modern cities, public places are an important guarantee for the safe operation of cities. While it brings convenience to people's lives, it also hides hidden dangers that cause public health risks. The dense flow of people in public places, suitable temperature and humidity environment, rich sources of nutrients, and complex equipment systems create conditions for the growth and reproduction of microorganisms. Taking urban rail transit places as an example, rail transit and stations have built underground spaces focusing on transfers and public activities. This space integrates multiple functions such as commuting, entertainment, catering, and leisure. The complex and multifunctional spatial layout accelerates the connection between urban areas and increases the spread and diffusion speed of pathogenic microorganisms, making public places a high-risk place environment for the lurking and transfer of pathogenic microorganisms. Moreover, once public health events occur in such key places, there are often problems of diversity, complexity, amplification, frequency, vulnerability, and difficulty in prevention. Risk identification and monitoring and early warning of pathogenic microorganisms in key places have always been an important proposition for the reconstruction of urban living environment and the construction of smart cities.
[0003] With the rapid development of biotechnology and information technology, the research, analysis and application of bioinformatics big data have become the forefront of scientific and technological development in various countries. Among them, high-throughput gene sequencing technology (Next-generation sequencing, NGS) represented by metagenomic sequencing has been widely used in the characteristic research of pathogenic microorganisms in the environment, biology and medicine by fully understanding the characteristics of pathogenic microorganism communities. Although the emergence of NGS technology is conducive to allowing humans to fully grasp the information of microbial communities in various environments, the excessive redundant gene information generated after sequencing has also brought difficulties to researchers in conducting bioinformatics data analysis and pathogenic microorganism monitoring and early warning. In particular, how to identify and discover key pathogenic microorganisms in massive sequencing data is the focus of researchers' current and future attention.
[0004] Different from traditional statistical methods, machine learning provides a method for analyzing complex relationships between massive data. For example, the Random Forest model (RF) can obtain key important variables through accuracy and computing power based on the decision tree (ntree). Compared with a single decision tree, the RF model introduces randomness and uses the bootstrap method to reduce bias and ensure the stability of the analysis results. The RF model is combined with the ten-fold cross-validation technology to further improve the sensitivity and specificity of indicator screening.
[0005] The use of machine learning technology to screen biomarkers has been applied in the diagnosis and treatment of clinical diseases. However, the identification of environmental pathogens in public places is different from that in clinical practice. Pathogenic microorganisms in the environment can be grouped according to time, location, medium, etc. At present, there is no case of identifying key environmental pathogens in public places based on machine learning technology, nor has the target microorganisms required for public health monitoring been established based on the seasonal characteristics of pathogens in public places. Summary of the invention
[0006] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art and provide a method, device, processor and computer-readable storage medium thereof for identifying, screening and processing key microorganisms in public places based on machine learning, which are simple to operate, have high analysis efficiency and a wide range of applications.
[0007] In order to achieve the above objectives, the method, device, processor and computer-readable storage medium thereof for realizing key microorganism identification, screening and processing for public place environment based on machine learning of the present invention are as follows:
[0008] The method for identifying, screening and treating key microorganisms in public places based on machine learning is mainly characterized in that the method comprises the following steps:
[0009] (1) Collect environmental microbial samples from public places in different seasons;
[0010] (2) Conduct high-throughput genome sequencing on samples;
[0011] (3) Cleaning and processing data sets;
[0012] (4) Establish a random forest model;
[0013] (5) Screening of key species.
[0014] The step (1) specifically comprises the following steps:
[0015] Collect environmental microbial samples from public places in different seasons; samples that are tested within 24 hours are stored in an environment of 2℃ to 8℃; samples that cannot be tested within 24 hours are stored at -80℃ or below.
[0016] Preferably, the step (2) specifically comprises the following steps:
[0017] (2.1) Conduct high-throughput gene sequencing on the collected samples;
[0018] (2.2) Extracting genomic DNA;
[0019] (2.3) Detection of DNA integrity;
[0020] (2.4) Perform data processing.
[0021] Preferably, the step (2.3) is specifically as follows:
[0022] The integrity of the extracted total DNA was tested by 1% agarose gel electrophoresis, and the concentration was tested by ultra-micro spectrophotometer. If the test was qualified, it was stored in a -20°C environment.
[0023] Preferably, the step (2.4) is specifically as follows:
[0024] The raw data of high-throughput sequencing are quality controlled to obtain high-quality effective reads, species annotation is performed, and statistical information on effective reads of sequencing data is obtained to perform high-throughput analysis of the composition and abundance of microorganisms at the domain, kingdom, phylum, class, order, family, genus or species.
[0025] Preferably, the step (3) is specifically as follows:
[0026] Sequencing data at the seasonal classification level were selected, and the data were filtered according to the standards of single sample abundance>0 and total abundance>0.05%. According to the research purpose, the data were divided into two groups, namely experimental group G1 and experimental group G2.
[0027] Preferably, the step (4) specifically comprises the following steps:
[0028] (4.1) Perform data preprocessing;
[0029] (4.2) Conduct model training;
[0030] (4.3) Perform feature selection.
[0031] Preferably, the step (4.1) specifically comprises the following steps:
[0032] The species abundance data obtained from sequencing are converted into a format suitable for use by machine learning models through data cleaning and normalization steps.
[0033] Preferably, the step (4.2) specifically includes the following steps:
[0034] The species abundance data is used as the input variable and the grouping is used as the output variable, which is input into the random forest model for training. Through the loop logic, the error OBB is selected to change to the minimum value to select the optimal node ntree. The binary classification and multi-classification of the grid parameter adjustment method are used to screen and optimize the optimal number of variables mtry for the binary tree.
[0035] Preferably, the step (4.3) specifically comprises the following steps:
[0036] The random forest model results provide the average Gini coefficient and mean reduction precision, and rank the importance of each microbial species.
[0037] Preferably, the step (5) specifically comprises the following steps:
[0038] The error curve was drawn for the initially screened microbial species using the ten-fold cross-validation method, and the number of microbial sequences at the lowest error point was used as the key microbial indicator for the final screening.
[0039] The device for implementing critical microbial screening and processing for public place environments based on machine learning has the following main features:
[0040] a processor configured to execute computer-executable instructions;
[0041] A memory stores one or more computer executable instructions. When the computer executable instructions are executed by the processor, the various steps of the above-mentioned method for identifying, screening and processing key microorganisms in public place environments based on machine learning are implemented.
[0042] The main feature of the processor for realizing key microorganism identification, screening and processing in public place environments based on machine learning is that the processor is configured to execute computer-executable instructions. When the computer-executable instructions are executed by the processor, the various steps of the method for realizing key microorganism identification, screening and processing in public place environments based on machine learning are realized.
[0043] The main feature of this computer-readable storage medium is that a computer program is stored thereon, and the computer program can be executed by a processor to implement the various steps of the above-mentioned method for identifying, screening and processing key microorganisms in a public place environment based on machine learning.
[0044] The method, device, processor and computer-readable storage medium thereof for identifying, screening and processing key microorganisms in public place environments based on machine learning of the present invention are adopted. By deeply mining environmental microbial gene sequencing data, a method for identifying characteristic microorganisms in public place environments is established based on RF machine learning, and 10 characteristic microorganisms for seasonal monitoring of environmental microorganisms in rail transit places are screened out based on test data. This method can greatly improve the utilization value of sequenced data. The present invention can be used to identify and identify microorganisms that play a key role in different experimental groups, greatly improve the analytical utilization value and analysis efficiency of public place environmental gene sequencing data, discover key microorganisms in public place environments, and provide characteristic monitoring targets for pathogenic microorganism monitoring and early warning in the environment, which has extremely high application value and social significance. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] Figure 1 A flow chart of the method for realizing key microorganism identification, screening and processing in public place environments based on machine learning of the present invention.
[0046] Figure 2 A schematic diagram of the classification of surface microorganisms in different seasons according to the method for identifying, screening and treating key microorganisms in public place environments based on machine learning of the present invention.
[0047] Figure 3a , 3b 3c is a schematic diagram of the summer key microorganism screening results of the method for identifying, screening and treating key microorganisms in public place environments based on machine learning of the present invention.
[0048] Figure 4a , 4b 4c is a schematic diagram of the autumn key microorganism screening results of the method for identifying, screening and treating key microorganisms in public place environments based on machine learning of the present invention.
[0049] Figure 5a , 5b 5c is a schematic diagram of the winter key microorganism screening results of the method for identifying, screening and treating key microorganisms in a public place environment based on machine learning of the present invention. DETAILED DESCRIPTION
[0050] In order to more clearly describe the technical content of the present invention, further description is given below in conjunction with specific embodiments.
[0051] The method of the present invention for realizing the identification, screening and treatment of key microorganisms in a public place environment based on machine learning comprises the following steps:
[0052] (1) Collect environmental microbial samples from public places in different seasons;
[0053] (2) Conduct genome sequencing on the samples;
[0054] (3) Cleaning and processing data sets;
[0055] (4) Establish a random forest model;
[0056] (5) Screening of key species.
[0057] As a preferred embodiment of the present invention, the step (1) specifically comprises the following steps:
[0058] Collect environmental microbial samples from public places in different seasons; store samples that have been tested within 24 hours at 2℃ to 8℃; store samples that cannot be tested within 24 hours at -80℃ or below.
[0059] As a preferred embodiment of the present invention, the step (2) specifically comprises the following steps:
[0060] (2.1) Conduct high-throughput gene sequencing on the collected samples;
[0061] (2.2) Extracting genomic DNA;
[0062] (2.3) Detection of DNA integrity;
[0063] (2.4) Perform data processing.
[0064] As a preferred embodiment of the present invention, the step (2.3) is specifically as follows:
[0065] The integrity of the extracted total DNA was tested by 1% agarose gel electrophoresis, and the concentration was tested by ultra-micro spectrophotometer. If the test was qualified, it was stored in a -20°C environment.
[0066] As a preferred embodiment of the present invention, the step (2.4) is specifically as follows:
[0067] The raw data of high-throughput sequencing are quality controlled to obtain high-quality effective reads, species annotation is performed, and statistical information on effective reads of sequencing data is obtained to perform high-throughput analysis of the composition and abundance of microorganisms at the domain, kingdom, phylum, class, order, family, genus or species.
[0068] As a preferred embodiment of the present invention, the step (3) is specifically as follows:
[0069] Sequencing data at the seasonal classification level were selected, and the data were filtered according to the standards of single sample abundance>0 and total abundance>0.05%. According to the research purpose, the data were divided into two groups, namely experimental group G1 and experimental group G2.
[0070] As a preferred embodiment of the present invention, the step (4) specifically comprises the following steps:
[0071] (4.1) Perform data preprocessing;
[0072] (4.2) Conduct model training;
[0073] (4.3) Perform feature selection.
[0074] As a preferred embodiment of the present invention, the step (4.1) specifically includes the following steps:
[0075] The species abundance data obtained from sequencing are converted into a format suitable for use by machine learning models through data cleaning and normalization steps.
[0076] As a preferred embodiment of the present invention, the step (4.2) specifically includes the following steps:
[0077] The species abundance data is used as the input variable and the grouping is used as the output variable, which is input into the random forest model for training. Through the loop logic, the error OBB is selected to change to the minimum value to select the optimal node ntree. The binary classification and multi-classification of the grid parameter adjustment method are used to screen and optimize the optimal number of variables mtry for the binary tree.
[0078] As a preferred embodiment of the present invention, the step (4.3) specifically includes the following steps:
[0079] The random forest model results provide the average Gini coefficient and mean reduction precision, and rank the importance of each microbial species.
[0080] As a preferred embodiment of the present invention, the step (5) specifically comprises the following steps:
[0081] The error curve was drawn for the initially screened microbial species using the ten-fold cross-validation method, and the number of microbial sequences at the lowest error point was used as the key microbial indicator for the final screening.
[0082] The device for implementing critical microorganism screening in a public place environment based on machine learning of the present invention comprises:
[0083] a processor configured to execute computer executable instructions;
[0084] A memory stores one or more computer executable instructions. When the computer executable instructions are executed by the processor, the various steps of the above-mentioned method for realizing critical microorganism screening and processing for public place environments based on machine learning are implemented.
[0085] The processor of the present invention is used to implement key microorganism screening and processing for public place environments based on machine learning, wherein the processor is configured to execute computer-executable instructions. When the computer-executable instructions are executed by the processor, the various steps of the above-mentioned method for implementing key microorganism screening and processing for public place environments based on machine learning are implemented.
[0086] The computer-readable storage medium of the present invention stores a computer program thereon, and the computer program can be executed by a processor to implement the various steps of the above-mentioned method for screening and processing key microorganisms in public place environments based on machine learning.
[0087] The present invention establishes a method for screening and applying key microorganisms in the public place environment based on machine learning, which can be used for the screening, evaluation, monitoring and early warning of environmental microorganisms in public places. Through the machine learning random forest model, seasonal characteristic microorganisms on the surface of environmental facilities and equipment in rail transit places are discovered.
[0088] The primary purpose of the present invention is to provide a screening and evaluation method for iconic microorganisms in public place environments based on machine learning technology, establish target microorganisms for seasonal monitoring, mainly by applying random forest classification technology and ten-fold cross validation, by debugging the random classification decision tree (ntree) and the number of variables randomly sampled (mtry) on the test set data, and obtaining the importance ranking of different microbial factors according to the average accuracy and computing power (Gini, MDA) after model fitting; using ten-fold cross validation technology, evaluate the number and category of key microbial factors, and screen out key microorganisms. The technology of the present invention is practiced in the monitoring of environmental surface microorganisms in rail transit places, and 10 key pathogenic microorganisms in the seasonal classification of environmental microorganisms in rail transit places in summer, autumn and winter are identified.
[0089] The present invention comprises the following specific steps:
[0090] 1. Collect environmental microbial samples from public places in different seasons:
[0091] Environmental microbial samples (including air, surfaces, dust, water samples, etc.) should be collected from public places in different seasons. The samples should be tested as soon as possible after collection. Samples that can be tested within 24 hours can be stored in an environment of 2℃ to 8℃; samples that cannot be tested within 24 hours should be stored at -80℃ or below.
[0092] 2. Conduct high-throughput gene sequencing on samples:
[0093] (1) Conduct high-throughput gene sequencing (including metagenomic, 16S rRNA, ITS sequencing, etc.) on the collected samples;
[0094] (2) Genome extraction: Use a commercially available DNA extraction kit to extract the genomic DNA of the microorganism according to the manufacturer's instructions.
[0095] (3) DNA integrity test: The integrity of the extracted total DNA was tested by 1% agarose gel electrophoresis and the concentration was tested by ultra-micro spectrophotometer. If the test was qualified (DNA mass concentration ≥ 30 ng·μL-1, DNA mass ≥ 3μg), it was stored at -20°C for subsequent high-throughput sequencing. Currently, common sequencing platforms include Illumina, T7, etc., and sequencing technology has been commercialized.
[0096] (4) Data processing: The raw data of high-throughput sequencing were subjected to quality control to obtain high-quality effective reads. MetaPhlAn2 software was used for species annotation and statistics of effective read information of sequencing data, including total read length, total base number and average read length, to perform high-throughput analysis of the composition and abundance of microorganisms in domain, kingdom, phylum, class, order, family, genus or species.
[0097] 3. Clean and process the data set:
[0098] Sequencing data at the seasonal classification level were selected, and the data were filtered according to the standards of single sample abundance>0 and total abundance>0.05%. According to the research purpose, the data were divided into two groups, experimental group G1 and group G2.
[0099] 4. Build a random forest model:
[0100] Random forest is an ensemble learning method that creates and combines multiple decision trees to make predictions. The training set data is incorporated into the random forest model, and the decision trees in the random forest (ntree) and the optimal number of variables for random sampling (mtry) are tuned.
[0101] (1) Data preprocessing: The species abundance data obtained from sequencing is converted into a format suitable for use by machine learning models through steps such as data cleaning and normalization.
[0102] (2) Model training: The species abundance data is used as the input variable and the grouping is used as the output variable, and the random forest model is input for training. Through the loop logic, the error OBB is selected to change to the minimum value to select the optimal node ntree. The binary classification and multi-classification of the grid parameter adjustment method are used to screen and optimize the optimal number of variables mtry for the binary tree.
[0103] (3) Feature selection: The random forest model results provide the average Gini coefficient (Gini) and mean decrease accuracy (MDA), and rank the importance of each microbial species. The higher these scores, the more important they are in the group.
[0104] 5. Key species screening:
[0105] The error curve was drawn for the initially screened microbial species using the ten-fold cross-validation method, and the number of microbial sequences at the lowest error point was used as the key microbial indicator for the final screening.
[0106] The primary purpose of the present invention is to provide a screening and evaluation method for characteristic microorganisms in public place environments based on machine learning technology, and to establish key microorganisms for seasonal monitoring of rail transit environment.
[0107] The embodiment of the present invention adopts seasonal key microorganism screening and identification on the surface of rail transit environment facilities and equipment, including the following steps:
[0108] 1. Collect environmental microbial samples from public places in different seasons:
[0109] Two representative rail transit stations were selected to collect 18 ambient air aerosol samples in summer, autumn and winter. The sampler (sampling flow rate>300L / min) and the instrument collection port should be 1.0m to 1.5m above the ground and no less than 1.0m away from walls or obstacles. The sample collection time was no less than 30min, and the collection solution was PBS solution. Samples should be tested as soon as possible after collection. Samples that can be tested within 24h can be stored in an environment of 2℃ to 8℃; samples that cannot be tested within 24h should be stored at -80℃ or below.
[0110] 2. Conduct high-throughput gene sequencing on samples:
[0111] (1) Conduct high-throughput metagenomic sequencing on collected aerosol samples;
[0112] (2) Extraction of genomic DNA: Use a commercial DNA extraction kit to extract the genomic DNA of microorganisms according to the manufacturer's instructions.
[0113] (3) Detection of DNA integrity: The integrity of the extracted total DNA was detected by 1% agarose gel electrophoresis and the concentration was detected by ultra-micro spectrophotometer. After passing the test (DNA mass concentration ≥ 30 ng·μL-1, DNA mass ≥ 3μg), it was stored at -20°C and used for sequencing on the Illumina high-throughput sequencing platform.
[0114] (4) Data processing: The raw data of high-throughput sequencing were subjected to quality control to obtain high-quality effective reads. MetaPhlAn2 software was used for species annotation and statistics of effective read information of sequencing data, including total read length, total base number and average read length, to perform high-throughput analysis of the composition and abundance of microorganisms in domain, kingdom, phylum, class, order, family, genus or species.
[0115] 3. Selection of training data set:
[0116] Sequencing data at the seasonal classification level were selected, and the data were filtered according to the standards of single sample abundance>0 and total abundance>5%. The data were divided into three groups according to the season: Warm, Transmit, and Cold.
[0117] 4. Random forest model establishment:
[0118] (1) Data preprocessing: Sequencing data at the seasonal classification level were selected, and the data were filtered according to the criteria of single sample abundance > 0 and total abundance > 0.05%. The data were divided into two groups according to the research purpose, experimental group G1 and group G2. The species abundance data obtained from sequencing were converted into a format suitable for machine learning models through data cleaning, normalization and other steps.
[0119] (2) Model training: The species abundance data is used as the input variable and the seasonal grouping is used as the output variable, and the random forest model is input for training. Through the loop logic, the error OBB is selected to change to the minimum value to select the optimal node ntree. The binary classification and multi-classification of the grid parameter adjustment method are used to screen and optimize the optimal number of variables mtry for the binary tree.
[0120] (3) Feature selection: The results of the random forest model can provide the average Gini coefficient (Gini) and mean decrease accuracy (MDA), and rank the importance of each microbial species. The higher these scores are, the more important they are in the seasonal grouping.
[0121] 5. Key species screening:
[0122] The error curve was drawn for the initially screened microbial species using the ten-fold cross-validation method, and the number of microbial sequences at the lowest error point was used as the key microbial indicator for the final screening.
[0123] Table 1 Seasonal important microbial species on the surface of rail transit facilities and equipment
[0124]
[0125] The present invention provides a screening and evaluation method for characteristic microorganisms in public place environments based on machine learning. Compared with the prior art, the method has the following beneficial effects:
[0126] As important infrastructure of modern cities, public places are an important guarantee for the normal operation of urban order. As the main places for people's daily life, work, leisure and entertainment, public places have potential risks of microbial contamination during construction and use. Once a public health incident occurs, there are problems of complexity, amplification and difficulty in prevention. Therefore, the identification of key microorganisms is an important proposition in the construction of a public health system in urban public places.
[0127] Traditional microbial identification and testing in public places mainly uses artificial culture methods to analyze common pathogenic microorganisms such as bacteria, fungi or Legionella pneumophila. Although traditional culture methods can isolate active microorganisms, they are time-consuming, labor-intensive, and have single information results. They cover up important information about microbial species and cannot track the composition and changing characteristics of microbial species communities in a timely and efficient manner. They cannot meet the needs of the government and relevant management departments to identify, screen, and warn of risks of pathogenic microorganisms. High-throughput sequencing technology makes it possible to understand the community structure of pathogenic microorganisms in public places.
[0128] The random forest model is a statistical learning method that uses self-service resampling technology to extract multiple samples from the original sample, recombines them to generate a new training sample set to train the decision tree, and then establishes a decision tree "forest" to determine the final prediction result after "voting". This method belongs to nonlinear data processing and is based on ensemble learning technology. It can quickly process high-dimensional data and effectively prevent overfitting. After training the data, this method can also screen important features.
[0129] In the massive sequencing data, the identification of key indicator microorganisms is crucial. The combination of random forest algorithm and ten-fold cross validation technology can effectively solve the technical difficulties of key microorganism screening in complex sequencing data.
[0130] The present invention is formed based on the training of public place gene detection test data, taking into account the complexity of microorganisms in public place environments, and is more in line with the actual situation of microorganisms in public place environments. It can accurately and quickly predict the types and importance of indicator microorganisms under experimental conditions; this method uses the screened microbial species that affect the experimental grouping. It solves the problem of identifying key indicator pathogenic microorganisms in public place environments and realizes the prediction of pathogenic microorganisms in public places, which is of great significance for the identification and monitoring and early warning of pathogenic microorganisms in public places.
[0131] The specific implementation scheme of this embodiment can refer to the relevant description in the above embodiment, which will not be repeated here.
[0132] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.
[0133] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.
[0134] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.
[0135] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution device. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0136] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the corresponding program may be stored in a computer-readable storage medium, which, when executed, includes one of the steps of the method embodiment or a combination thereof.
[0137] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.
[0138] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.
[0139] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.
[0140] The method, device, processor and computer-readable storage medium for identifying, screening and processing key microorganisms in public place environments based on machine learning of the present invention are adopted. By deeply mining the environmental microbial gene sequencing data, a method for evaluating the iconic microorganisms in public place environments is established based on RF machine learning. Based on the test data, 10 characteristic microorganisms for seasonal monitoring of environmental microorganisms in rail transit places are screened out. This method can greatly improve the utilization value of sequenced data. The present invention can be used to identify and identify microorganisms that play a key role in different experimental groups, greatly improve the analytical utilization value and analysis efficiency of public place environmental gene sequencing data, discover key microorganisms in public place environments, and provide characteristic monitoring targets for pathogenic microorganism monitoring and early warning in the environment, which has extremely high application value and social significance.
[0141] In this specification, the present invention has been described with reference to specific embodiments thereof. However, it is apparent that various modifications and variations may be made without departing from the spirit and scope of the present invention. Therefore, the specification and drawings should be regarded as illustrative rather than restrictive.
Claims
1. A method for screening and processing key microorganisms in public place environments based on machine learning, characterized in that: The method comprises the following steps: (1) Collect environmental microbial samples from public places in different seasons; (2) Conduct high-throughput genome sequencing on samples; (3) Cleaning and processing data sets; (4) Establish a random forest model; (5) Screening of key species.
2. The method for screening and processing critical microorganisms in public place environments based on machine learning according to claim 1 is characterized in that: The step (1) specifically comprises the following steps: Collect environmental microbial samples from public places in different seasons; samples that are tested within 24 hours are stored in an environment of 2℃ to 8℃; samples that cannot be tested within 24 hours are stored at -80℃ or below.
3. The method for screening and processing critical microorganisms in public place environments based on machine learning according to claim 1, characterized in that: The step (2) specifically comprises the following steps: (2.1) Conduct high-throughput gene sequencing on the collected samples; (2.2) Extracting genomic DNA; (2.3) Detection of DNA integrity; (2.4) Perform data processing.
4. The method for screening and processing critical microorganisms in public place environments based on machine learning according to claim 3 is characterized in that: The step (2.3) is specifically: The integrity of the extracted total DNA was tested by 1% agarose gel electrophoresis, and the concentration was tested by ultra-micro spectrophotometer. If the test was qualified, it was stored in a -20°C environment.
5. The method for screening and processing critical microorganisms in public place environments based on machine learning according to claim 3 is characterized in that: The step (2.4) is specifically: The raw data of high-throughput sequencing are quality controlled to obtain high-quality effective reads, species annotation is performed, and statistical information on effective reads of sequencing data is obtained to perform high-throughput analysis of the composition and abundance of microorganisms at the domain, kingdom, phylum, class, order, family, genus or species.
6. The method for screening and processing critical microorganisms in public place environments based on machine learning according to claim 1, characterized in that: The step (3) is specifically as follows: Sequencing data at the seasonal classification level were selected, and the data were filtered according to the standards of single sample abundance>0 and total abundance>0.05%. According to the research purpose, the data were divided into two groups, namely experimental group G1 and experimental group G2.
7. The method for screening and treating critical microorganisms in public place environments based on machine learning according to claim 1, characterized in that: The step (4) specifically comprises the following steps: (4.1) Perform data preprocessing; (4.2) Perform model training; (4.3) Perform feature selection.
8. The method for screening and treating critical microorganisms in public place environments based on machine learning according to claim 7, characterized in that: The step (4.1) specifically comprises the following steps: The species abundance data obtained from sequencing are converted into a format suitable for use by machine learning models through data cleaning and normalization steps.
9. The method for screening and processing critical microorganisms in public place environments based on machine learning according to claim 7, characterized in that: The step (4.2) specifically comprises the following steps: The species abundance data is used as the input variable and the grouping is used as the output variable, which is input into the random forest model for training. Through the loop logic, the error OBB is selected to change to the minimum value to select the optimal node ntree. The binary classification and multi-classification of the grid parameter adjustment method are used to screen and optimize the optimal number of variables mtry for the binary tree.
10. The method for screening and processing critical microorganisms in public place environments based on machine learning according to claim 7, characterized in that: The step (4.3) specifically comprises the following steps: The random forest model results provide the average Gini coefficient and mean reduction precision, and rank the importance of each microbial species.
11. The method for screening and processing critical microorganisms in public place environments based on machine learning according to claim 1, characterized in that: The step (5) specifically comprises the following steps: The error curve was drawn for the initially screened microbial species using the ten-fold cross-validation method, and the number of microbial sequences at the lowest error point was used as the key microbial indicator for the final screening.
12. A device for implementing critical microbial screening in public places based on machine learning, characterized in that: The device comprises: a processor configured to execute computer-executable instructions; A memory storing one or more computer executable instructions, which, when executed by the processor, implements the various steps of the method for screening and processing critical microorganisms in public place environments based on machine learning as described in any one of claims 1 to 11.
13. A processor for implementing critical microbial screening for public place environments based on machine learning, characterized in that: The processor is configured to execute computer-executable instructions. When the computer-executable instructions are executed by the processor, the various steps of the method for screening and processing critical microorganisms in public place environments based on machine learning described in any one of claims 1 to 11 are implemented.
14. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program can be executed by a processor to implement the various steps of the method for screening and processing critical microorganisms in public place environments based on machine learning as described in any one of claims 1 to 11.
Citation Information
Patent Citations
Method for discriminating and screening key microorganisms in Noni natural fermentation process
CN109112225A
