A method for analyzing microbial community clustering

Abstract

The present invention relates to the field of agricultural environment, and in particular to a method for analyzing microbial community clustering. The method for analyzing microbial community clustering of the present invention comprises the following steps: first, based on the absolute abundance of different individuals in the microbial community that changes over time, the relative change rate of the individual is calculated to describe the change in the growth trend of the microorganism; second, considering the similarity of the growth trends between different individuals, the correlation coefficient of the relative change rate is calculated, and a similarity matrix is ​​constructed based on the correlation coefficient of the relative change rate; finally, affinity propagation clustering analysis is performed based on the obtained similarity matrix to obtain cluster centers. By constructing a similarity matrix based on the relative change rate correlation coefficient and using affinity propagation clustering to calculate the cluster centers, while taking into account the competition and symbiotic relationships between individuals, more effective cluster centers and cluster clusters for the study of microbial community relationships can be obtained.

Description

Technical Field

[0001] The present invention relates to the field of agricultural environment, and in particular to an analysis method for clustering of microbial communities. Background Art

[0002] Livestock and poultry manure is rich in nitrogen and phosphorus and is often used as fertilizer to improve agricultural production efficiency. However, improper manure handling leads to nitrogen loss, which not only reduces the agricultural value of the fertilizer but also causes environmental pollution. Aerobic composting is a green and effective method for treating livestock and poultry manure. It transforms solid organic matter into stable humus through microbial activity, achieving mineralization and humification of organic matter. Compost can also serve as a soil conditioner, improving soil fertility, promoting plant growth, and protecting the environment. Aerobic composting involves multiple physical and chemical reactions as well as complex microbial community succession and interspecies competition. A comprehensive understanding and characterization of the microbial interactions during composting can help improve the reproducibility and effectiveness of compost products. Screening a large number of microorganisms within the microbial community and identifying those with specific interactions can effectively narrow the scope of experimental research and significantly reduce experimental costs. It also provides opportunities for further research on interspecies competition within microorganisms during composting.

[0003] In addition, there are a large number of microbial communities colonizing the human intestine. The intestinal microbial community has evolved with the host and is an integral part of the human body. Normally structured intestinal flora plays an important role in host metabolism, immune system protection, and normal function of the nervous system. Studies have found that disruption of the normal structure of the intestinal flora is associated with a variety of gastrointestinal diseases and extraintestinal diseases (allergies, asthma, obesity, and cardiovascular diseases, etc.) as well as a series of mental illnesses. Related studies have found that homeostasis disorders of the intestinal flora can participate in the occurrence of various cancers, including lung cancer, by affecting the host's metabolism, causing inflammation, and acting on the immune system. Many studies have confirmed that changes in the structure of the intestinal flora are closely related to the occurrence of lung cancer. Analyzing the relationships among intestinal microbial communities will also help explore the intrinsic connection between changes in the structure of the intestinal flora and the occurrence of lung cancer.

[0004] Existing microbial clustering methods only consider symbiotic growth trends among microorganisms, grouping microorganisms with similar trends into a single cluster. Microorganisms within these clusters may have symbiotic relationships, sharing similar characteristics and functions. However, considering only symbiotic relationships is incomplete. To fully examine the interactions between microorganisms, clustering methods should also consider competitive relationships between microorganisms. Summary of the Invention

[0005] In order to overcome the above problems, the purpose of the present invention is to provide an analysis method for microbial community clustering, which constructs a similarity matrix based on the relative change rate correlation coefficient, calculates the cluster center using affinity propagation clustering, and takes into account the competition and symbiotic relationship between individuals, so as to obtain more effective cluster centers and cluster clusters for the study of microbial community relationships.

[0006] To achieve the above-mentioned object, the present invention provides a method for analyzing microbial community clustering, which is characterized by comprising the steps of:

[0007] (1) Collect microbial samples in the target biological environment in the order of culture time;

[0008] Screening microbial samples according to characteristic growth periods of microorganisms to obtain microbial analysis samples for research;

[0009] (2) Obtain the absolute abundance data of each microorganism in each microbial analysis sample based on high-throughput sequencing technology;

[0010] (3) Calculation of relative change rate sequence of microorganisms based on absolute abundance

[0011]

[0012] x i (k) is the absolute abundance of the i-th microorganism in the k-th sampling, v i (k) is the relative change rate of the i-th microorganism at the k-th sampling time; where min(x i k()x, i k(-1 represents x i (k) and x i The minimum value among (k-1);

[0013] (4) According to the relative change rate sequence of microorganisms, the relative change rate correlation coefficient is calculated according to formula (2):

[0014]

[0015] where corr ij The correlation coefficient of the relative change rate between the i-th microorganism and the j-th microorganism, Cov(x i ,x j ) represents the covariance of the relative change rate between the i-th microorganism and the j-th microorganism; Var(x i ) represents the variance of the relative change rate of the i-th microorganism, Var(x j ) represents the variance of the relative change rate of the jth microorganism;

[0016] (5) Construct similarity matrix:

[0017] The competitive similarity s between the i-th microorganism and the j-th microorganism is calculated as formula (3): ij , the competitive similarity between all individuals and themselves is replaced by the median of the remaining similarities, and finally the similarity matrix of the microbial community at the genus level is obtained;

[0018]

[0019] (6) Obtain affinity propagation clustering graph:

[0020] Based on the similarity matrix, the affinity propagation cluster analysis method is used to obtain the cluster centers. Individuals with the same cluster centers are grouped into one cluster, and the affinity propagation cluster diagram is drawn according to the cluster centers and cluster data.

[0021] As a preferred embodiment, in step (1), the characteristic growth periods of the microorganisms include a temperature rising period, a high temperature period, a temperature falling period, and a mature period. All microbial samples are classified according to the temperature rising period, the high temperature period, the temperature falling period, and the mature period, and 1 to 3 samples are selected from each period to obtain k microbial analysis samples for research, where 4 < k < 12.

[0022] As a preferred embodiment, in step (2), the 16S rRNA gene is used to perform Illumina NovaSeq high-throughput sequencing technology on k microbial analysis samples, and the absolute abundance of m microorganisms is detected in each microbial analysis sample. Microorganisms with an absolute abundance of less than 100 in each microbial analysis sample are removed, and finally, the absolute abundance data of i microorganisms are screened.

[0023] As a preferred solution, in step (6), the specific process of obtaining the affinity propagation clustering graph is:

[0024] Based on the similarity matrix, two matrices with all zeros are first constructed as the initial attribution matrix and the initial attraction matrix, and the value of each cell in the attraction matrix is ​​calculated using formula (4):

[0025]

[0026] where r ik is the attraction between the i-th microorganism and the k-th microorganism, a ik is the degree of belonging between the i-th microorganism and the k-th microorganism, i and k represent the row and column indexes in the corresponding matrix respectively, and k′ represents the column index that is not equal to k. The degree of belonging matrix is ​​updated as shown in formula (5):

[0027]

[0028] Where i′ represents a row index that is not equal to k, and i″ represents a row index that is not equal to both i and k.

[0029] Repeat the above steps to update the attraction matrix and the attribution matrix until the specified number of iterations is reached. Finally, the attraction matrix and the attribution matrix are added together to obtain the evaluation matrix. The microorganisms corresponding to the maximum values ​​in each row of the evaluation matrix are selected as cluster centers, and individuals with the same cluster center are grouped into a cluster. An affinity propagation cluster diagram is drawn based on the cluster centers and cluster cluster data.

[0030] As a preferred embodiment, the target biological environment in step (1) is aerobic compost or human intestine.

[0031] The design concept of this invention is as follows: First, based on the absolute abundance changes of different individuals in the microbial community over time, the relative change rate of each individual is calculated to describe the changes in microbial growth trends. Second, considering the similarity of growth trends between different individuals, the correlation coefficient of the relative change rate is calculated. Based on the correlation coefficient of the relative change rate, a similarity matrix is ​​constructed. Finally, affinity propagation clustering analysis is performed on the obtained similarity matrix to determine the cluster centers.

[0032] The advantages of this method are that it considers the changing patterns of microbial growth trends, measures the strength of relationships between microorganisms by calculating the relative rate of change correlation coefficient, constructs a similarity matrix based on the relative rate of change correlation coefficient, and uses affinity propagation clustering to calculate cluster centers and obtain an affinity propagation aggregation graph. Compared with traditional microbial community clustering methods, this method uses the relative rate of change correlation coefficient to measure the strength of relationships between microorganisms, while also considering the competition and symbiotic relationships between different microbial species, resulting in more effective cluster centers and clusters. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a heat map of the similarity of microbial communities in the pile at the genus level in the embodiment;

[0034] Figure 2 Schematic diagram of affinity propagation clustering in the embodiment. DETAILED DESCRIPTION

[0035] For a better understanding of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific examples.

[0036] Example 1 A method for analyzing microbial community clustering, comprising the steps of:

[0037] Step 1: Microbial compost cultivation and sampling

[0038] Take a composting experiment at a certain experimental chicken farm as an example. The composting experiment employed static forced ventilation fermentation, with the pile turned regularly according to the fermentation temperature, for a total of 40 days. The entire composting period is divided into four phases: the warming phase, the high-temperature phase, the cooling phase, and the mature phase. The warming phase is the period from the initial state until the pile temperature reaches 60°C (approximately Days 0-1 in this experiment); the high-temperature phase is the period when the pile temperature remains above 60°C (Days 1-8 in this experiment); the cooling phase is when the pile temperature drops below 60°C (Days 9-17 in this experiment); and the mature phase is when the temperature drops below 40°C (Days 18-38 in this experiment).

[0039] Step 2: Screen samples and obtain the absolute abundance of microorganisms in each sample

[0040] Compost samples from the composting heating period (day 1), high temperature period (days 3 and 4), cooling period (day 9), and maturity period (days 24 and 38) were selected for high-throughput sequencing. The 16S rRNA gene was sequenced using Illumina NovaSeq high-throughput sequencing technology to obtain the absolute abundance of microorganisms at different taxonomic levels at each time point. Based on the sequencing, the absolute abundances of more than 600 microorganisms on days 1, 3, 4, 9, 24, and 38 were obtained. Among them, the absolute abundances of many microorganisms were relatively low. To ensure the validity of the model, microorganisms with absolute abundances below 100 on days 1, 3, 4, 9, 24, and 38 were removed. Finally, a total of 45 microorganisms were obtained for study, as shown in Table 1. Due to the large amount of data, only the absolute abundance values ​​of 10 microorganisms are listed.

[0041] Table 1 Absolute abundance values ​​of some microorganisms

[0042]

[0043]

[0044] Step 3: Calculate the relative change rate sequence of 45 microorganisms

[0045] Let the absolute abundance of the i-th microorganism in the k-th sampling be x i (k), i=1,2,...,45,k=1,2,...,6. Then the relative change rate v of the i-th microorganism at the k-th sampling i (k) As shown in formula (1):

[0046]

[0047] Among them minx i (k(x i )k,-(represents x i (k) and x iThe minimum value in (k-1), where k=1, 2, 3, 4, 5, 6 represents the 1st day, 3rd day, 4th day, 9th day, 24th day, and 38th day.

[0048] Step 4: Based on the relative change rate sequences of the 45 microorganisms obtained, calculate the relative change rate correlation coefficient according to formula (2):

[0049]

[0050] where corr ij The correlation coefficient of the relative change rate between the i-th microorganism and the j-th microorganism, Cov(x i ,x j ) represents the covariance of the relative change rate between the i-th microorganism and the j-th microorganism; Var(x i )Var(x j ) represent the variance of the relative change rate of the i-th microorganism and the j-th microorganism, respectively.

[0051] Step 5: Construct similarity matrix:

[0052] Normalize the relative change rate correlation coefficient and take a negative value, as shown in formula (3), to calculate the competition similarity s between the i-th microorganism and the j-th microorganism ij In addition, the competitive similarity between all individuals and themselves is replaced by the median of the other similarities, and finally the similarity matrix of the microbial community at the genus level is obtained, as shown in Table 2. Due to the large amount of data, Table 2 only lists the competitive similarity data of 6 microorganisms. The matrix is ​​presented in the form of a heat map as follows Figure 1 shown.

[0053]

[0054] Table 2 Competitive similarity s of 6 microorganisms ij

[0055] 0 1 2 3 4 5 6 0 0 -0.04451 -0.66333 -0.09926 -0.03384 -0.44709 -0.03701 1 -0.04451 0 -0.69613 -0.24199 -0.13414 -0.29704 -0.13848 2 -0.66333 -0.69613 0 -0.43594 -0.71691 -0.69835 -0.72565 3 -0.09926 -0.24199 -0.43594 0 -0.08273 -0.69239 -0.08722 4 -0.03384 -0.13414 -0.71691 -0.08273 0 -0.55619 -0.00031 5 -0.44709 -0.29704 -0.69835 -0.69239 -0.55619 0 -0.54908 6 -0.03701 -0.13848 -0.72565 -0.08722 -0.00031 -0.54908 0

[0056] Step 6: Construct affinity propagation aggregation graph:

[0057] Based on the similarity matrix, we first construct two matrices with all zeros as the initial attribution matrix and the initial attraction matrix. Then, we use formula (4) to calculate the value of each cell in the attraction matrix:

[0058]

[0059] where r ik is the attraction between the i-th microorganism and the k-th microorganism, a ik is the degree of belonging between the i-th microorganism and the k-th microorganism, s ikis the competitive similarity between the i-th microorganism and the k-th microorganism, i and k represent the row and column indexes in the corresponding matrix respectively, and k′ represents the column index that is not equal to k. The attribution matrix is ​​updated as shown in formula (5):

[0060]

[0061] a kk Indicates a ik In the case of i=k, r kk Represents r ik In the case where i=k, i′ represents a row index that is not equal to k, and i″ represents a row index that is not equal to both i and k.

[0062] Repeat the above steps to update the attraction matrix and the attribution matrix until the specified number of iterations. Finally, the attraction matrix and the attribution matrix are added to obtain the evaluation matrix, as shown in Table 3. Due to the large amount of data, only part of the data in the evaluation matrix is ​​listed. The microorganism corresponding to the maximum value of each row in the evaluation matrix is ​​selected as the cluster center, and individuals with the same cluster center are clustered into a cluster. The clustering results are shown in Table 3. Figure 2 shown.

[0063] Table 3 Evaluation matrix

[0064]

[0065]

[0066] The above-described embodiment merely represents one embodiment of the present invention. While the description is relatively specific and detailed, it should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art may make various modifications and improvements without departing from the spirit of the present invention, and these modifications and improvements fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A method for analyzing microbial community clustering, characterized in that: Including steps: (1) Collect microbial samples in the target biological environment in the order of culture time; Screening microbial samples according to characteristic growth periods of microorganisms to obtain microbial analysis samples for research; (2) Obtain the absolute abundance data of each microorganism in each microbial analysis sample based on high-throughput sequencing technology; (3) Based on the absolute abundance data, the relative change rate of microorganisms was calculated according to formula (1): x i (k) is the absolute abundance of the i-th microorganism in the k-th sampling, v i (k) is the relative change rate of the i-th microorganism at the k-th sampling time; where min(x i (k),x i (k-1)) represents x i (k) and x i The minimum value among (k-1); (4) According to the relative change rate of microorganisms, the correlation coefficient of the relative change rate is calculated according to formula (2): where corr ij The correlation coefficient of the relative change rate between the i-th microorganism and the j-th microorganism, Cov(x i ,x j ) represents the covariance of the relative change rate between the i-th microorganism and the j-th microorganism; Var(x i ) represents the variance of the relative change rate of the i-th microorganism, Var(x j ) represents the variance of the relative change rate of the jth microorganism; (5) Construct similarity matrix: Based on the correlation coefficient of the relative change rate, the competitive similarity s between the i-th microorganism and the j-th microorganism is calculated as follows: ij , the competitive similarity between all individuals and themselves is replaced by the median of the remaining similarities, and finally the similarity matrix of the microbial community at the genus level is obtained; (6) Obtain affinity propagation clustering graph: Based on the similarity matrix, the affinity propagation cluster analysis method is used to obtain the cluster centers. Individuals with the same cluster centers are grouped into one cluster, and the affinity propagation cluster diagram is drawn according to the cluster centers and cluster data.

2. The method for analyzing microbial community clustering according to claim 1, characterized in that: In the step (1), the characteristic growth periods of the microorganisms include a temperature rise period, a high temperature period, a temperature drop period, and a mature period; all microbial samples are classified according to the temperature rise period, the high temperature period, the temperature drop period, and the mature period, and 1 to 3 samples are selected from each period to obtain k microbial analysis samples for research, wherein 4<k<12.

3. The method for analyzing microbial community clustering according to claim 2, characterized in that: In the step (2), the 16S rRNA gene is used to perform Illumina NovaSeq high-throughput sequencing technology on k microbial analysis samples, and the absolute abundance of m microorganisms is detected in each microbial analysis sample. Microorganisms with an absolute abundance of less than 100 in each microbial analysis sample are removed, and finally, the absolute abundance data of i microorganisms are obtained by screening.

4. The method for analyzing microbial community clustering according to claim 3, wherein: In step (6), the specific process of obtaining the affinity propagation clustering graph is: Based on the similarity matrix, two matrices with all zeros are first constructed as the initial attribution matrix and the initial attraction matrix, and the value of each cell in the attraction matrix is ​​calculated using formula (4): where r ik is the attraction between the i-th microorganism and the k-th microorganism, a ik is the degree of belonging between the i-th microorganism and the k-th microorganism, s ik is the competitive similarity between the i-th microorganism and the k-th microorganism, i and k represent the row and column indexes in the corresponding matrix respectively, and k′ represents the column index that is not equal to k. The attribution matrix is ​​updated as shown in formula (5): a kk Indicates a ik In the case of i=k, r kk Represents r ik In the case where i=k, i′ represents a row index that is not equal to k, and i″ represents a row index that is not equal to both i and k; Repeat the above steps to update the attraction matrix and the attribution matrix until the specified number of iterations is reached. Finally, the attraction matrix and the attribution matrix are added together to obtain the evaluation matrix. The microorganisms corresponding to the maximum values ​​in each row of the evaluation matrix are selected as cluster centers, and individuals with the same cluster center are grouped into a cluster. An affinity propagation cluster diagram is drawn based on the cluster centers and cluster cluster data.

5. The method for analyzing microbial community clustering according to any one of claims 1 to 4, characterized in that: In the step (1), the target biological environment is aerobic compost or human intestine.

Images