Biodiversity analysis method and system based on aquatic eDNA

Abstract

The invention relates to the technical field of biodiversity analysis, in particular to a biodiversity analysis method and system based on aquatic eDNA. The method comprises the following steps: associating eDNA data with water flow characteristics of a river; the biological diversity rapid analysis model is used for determining the eDNA data of the downstream river according to the eDNA data of the upstream river and the water flow characteristics from the upstream river to the downstream river; and combining the eDNA spatial distribution map with the biodiversity rapid analysis model to obtain a biodiversity rapid analysis map. The eDNA data is projected on the river distribution map of the water environment to obtain the eDNA spatial distribution map of the water environment, the eDNA distribution condition in the whole water environment can be mastered, then the eDNA spatial distribution map is combined with the biological diversity rapid analysis model to obtain the biological diversity rapid analysis map, rapid analysis of eDNA distribution can be achieved, and the analysis efficiency is improved. And finally, rapid analysis of biodiversity in the whole water environment is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of biodiversity analysis, and in particular to a method and system for biodiversity analysis based on aquatic eDNA. Background Art

[0002] Environmental DNA (eDNA) is a general term for mixed DNA released into the environment by various organisms. eDNA metabarcoding refers to a high-throughput multiple species identification method using DNA directly extracted from environmental samples (such as water, soil, etc.). This technology integrates DNA barcoding and high-throughput sequencing technology. It extracts DNA from environmental samples, uses specific primers for PCR amplification, and sequences the amplified products to obtain operational taxonomic units (OTUs) for species identification. Therefore, eDNA can be used to analyze biodiversity in aquatic environments.

[0003] When using eDNA to analyze biodiversity in aquatic environments in the prior art, it is usually necessary to extract eDNA data from each river in the aquatic environment one by one. Although the distribution of eDNA in the entire aquatic environment can be analyzed, the whole process is complicated and cumbersome, making it difficult to achieve rapid analysis of biodiversity. Summary of the invention

[0004] The purpose of the present invention is to provide a biodiversity analysis method and system based on aquatic eDNA to solve the technical problem that the existing technology needs to extract eDNA data from each river in the water environment one by one, the whole process is complicated and cumbersome, and it is difficult to achieve rapid analysis of biodiversity.

[0005] In order to solve the above technical problems, the present invention specifically provides the following technical solutions: A biodiversity analysis method based on aquatic eDNA comprises the following steps: Set up eDNA sampling points at the downstream of each river in the water environment and obtain eDNA data at each eDNA sampling point; Project the eDNA data onto the river distribution map of the water environment to obtain the eDNA spatial distribution map of the water environment; The eDNA data is associated with the flow characteristics of the river, and a rapid biodiversity analysis model is obtained for determining the eDNA data of the downstream river based on the eDNA data of the upstream river and the flow characteristics of the upstream river to the downstream river; The eDNA spatial distribution map is combined with the rapid biodiversity analysis model to obtain a rapid biodiversity analysis map.

[0006] As a preferred solution of the present invention, the method for obtaining eDNA data at an eDNA sampling point includes: Collect surface water samples at the eDNA sampling points to obtain water environment samples at the eDNA sampling points; The water environment sample was filtered through a mesh sieve, the solid matter was collected with a sterile centrifuge tube, and sodium chloride solution and benzalkonium chloride solution were added at a ratio of 1 g:1 mL:1 mL to obtain a solid sample. The filtered water sample was then collected in a sterile ziplock bag and emptied according to the ratio. Add solid benzalkonium chloride to obtain a liquid sample; Add proteinase K solution, sodium dodecyl sulfate solution and ethylenediaminetetraacetic acid solution to the solid sample and mix well. oscillation , centrifuging, retaining the supernatant, then mixing the supernatant with the liquid sample, filtering with a filter membrane, and intercepting the eDNA in the water sample onto the filter membrane; The eDNA on the filter membrane was extracted using an eDNA extraction kit, and the concentration and quality of the extracted genomic eDNA were tested. Then, a PE gene library was constructed, bridge PCR was performed, and finally, metagenomic sequencing was performed using the Illumina HiSeq high-throughput sequencing platform to obtain the eDNA data of the eDNA sampling point.

[0007] As a preferred embodiment of the present invention, the method for constructing the eDNA spatial distribution map of the water environment comprises: In the river distribution map of the water environment, the eDNA data is displayed at the corresponding river according to the eDNA sampling points to obtain the eDNA spatial distribution map of the water environment.

[0008] As a preferred embodiment of the present invention, the method for constructing the rapid biodiversity analysis model comprises: The eDNA data of the upstream rivers in the eDNA spatial distribution map and the water flow from the upstream rivers to the downstream rivers are used as the input of the neural network, and the eDNA data of the downstream rivers in the eDNA spatial distribution map are used as the output of the neural network; Using a neural network to map input and output to train, to obtain the biodiversity rapid analysis model; The biodiversity rapid analysis model is: Downstream-eDNA=BP(Upstream-eDNA, Q-UptoDown); In the formula, Downstream-eDNA is the predicted value of the eDNA data of the downstream river, Upstream-eDNA is the eDNA data of the upstream river, Q-UptoDown is the flow characteristics from the upstream river to the downstream river, and BP is a neural network; Among them, the loss function for training the rapid analysis model of biodiversity is: ; Where L is the loss function, (Downstream-eDNA)1 is the predicted value of the eDNA data of the downstream river in the first layer of the water environment, (Downstream-eDNA-real)1 is the true value of the eDNA data of the downstream river in the first layer of the water environment, and (Downstream-eDNA-real) k is the true value of the eDNA data of the downstream river in the kth layer of the water environment, (Downstream-eDNA) k-1 is the predicted value of the eDNA data of the downstream river in the k-1th layer of the water environment, (Q-UptoDown) k is the flow characteristics from the upstream river to the downstream river at the kth level in the water environment, and n is the total number of upstream and downstream levels of rivers in the water environment; BP((Downstream-eDNA) k-1 ,(Q-UptoDown) k ) is the predicted value of the eDNA data of the downstream river in the kth layer of the water environment; as well as All are L1 norm forms.

[0009] As a preferred embodiment of the present invention, a method for obtaining flow characteristics from an upstream river to a downstream river includes: During the collection period of eDNA data, the velocity-area method was used to obtain the flow rate series of the upstream river and the flow rate series of the downstream river at the sampling points of the upstream river and the downstream river; The mutual information is used to quantify the degree of dependence of the water flow sequence of the downstream river on the water flow sequence of the upstream river, and the degree of dependence of the water flow sequence of the downstream river on the water flow sequence of the upstream river is used as the water flow characteristic from the upstream river to the downstream river; The quantitative formula for the flow characteristics from the upstream river to the downstream river is: ; Where Q-UptoDown is the flow characteristic from the upstream river to the downstream river, is the flow sequence of the upstream river, is the flow sequence of the downstream river, for and The joint probability distribution of for The marginal probability distribution of for The marginal probability distribution of .

[0010] As a preferred embodiment of the present invention, the method for constructing the biodiversity rapid analysis map comprises: In the eDNA spatial distribution map of the water environment, the biodiversity rapid analysis model is encapsulated in the corresponding eDNA data to obtain the biodiversity rapid analysis map.

[0011] As a preferred embodiment of the present invention, the present invention provides a biodiversity analysis system based on aquatic eDNA, which is applied to the biodiversity analysis method based on aquatic eDNA, and the system comprises: The sampling unit is used to set up eDNA sampling points at the downstream of each river in the water environment and obtain eDNA data of each eDNA sampling point; A data processing unit, used to associate the eDNA data with the flow characteristics of the river, and obtain a biodiversity rapid analysis model for determining the eDNA data of the downstream river based on the eDNA data of the upstream river and the flow characteristics of the upstream river to the downstream river; A graphics construction unit is used to project the eDNA data onto a river distribution map of the water environment to obtain an eDNA spatial distribution map of the water environment; And, it is used to combine the eDNA spatial distribution map with the rapid analysis model of biodiversity to obtain a rapid analysis map of biodiversity.

[0012] As a preferred solution of the present invention, the biodiversity rapid analysis model is encapsulated into a biodiversity rapid analysis map through a selection button control, and the biodiversity rapid analysis model is activated by selecting the selection button control to acquire eDNA data.

[0013] As a preferred solution of the present invention, the eDNA data is displayed in a biodiversity rapid analysis map through a digital display frame control, and the digital display frame control is informationally interconnected with an output end of a biodiversity rapid analysis model.

[0014] As a preferred solution of the present invention, the method for constructing a rapid analysis model of biodiversity by a data processing unit includes: The eDNA data of the upstream rivers in the eDNA spatial distribution map and the water flow from the upstream rivers to the downstream rivers are used as the input of the neural network, and the eDNA data of the downstream rivers in the eDNA spatial distribution map are used as the output of the neural network; Using a neural network to map input and output to train, to obtain the biodiversity rapid analysis model; The biodiversity rapid analysis model is: Downstream-eDNA=BP(Upstream-eDNA, Q-UptoDown); In the formula, Downstream-eDNA is the predicted value of the eDNA data of the downstream river, Upstream-eDNA is the eDNA data of the upstream river, Q-UptoDown is the flow characteristics from the upstream river to the downstream river, and BP is a neural network; Among them, the loss function for training the rapid analysis model of biodiversity is: ; Where L is the loss function, (Downstream-eDNA)1 is the predicted value of the eDNA data of the downstream river in the first layer of the water environment, (Downstream-eDNA-real)1 is the true value of the eDNA data of the downstream river in the first layer of the water environment, and (Downstream-eDNA-real) k is the true value of the eDNA data of the downstream river in the kth layer of the water environment, (Downstream-eDNA) k-1 is the predicted value of the eDNA data of the downstream river in the k-1th layer of the water environment, (Q-UptoDown) k is the flow characteristics from the upstream river to the downstream river at the kth level in the water environment, and n is the total number of upstream and downstream levels of rivers in the water environment; BP((Downstream-eDNA) k-1 ,(Q-UptoDown) k ) is the predicted value of the eDNA data of the downstream river in the kth layer of the water environment; as well as All are L1 norm forms.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention projects eDNA data onto a river distribution map of a water environment to obtain an eDNA spatial distribution map of the water environment, which can grasp the eDNA distribution in the entire water environment. Then, by combining the eDNA spatial distribution map with a rapid biodiversity analysis model, a rapid biodiversity analysis map is obtained, which can realize rapid analysis of eDNA distribution and ultimately realize rapid analysis of biodiversity in the entire water environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0017] Figure 1 A flow chart of a biodiversity analysis method based on aquatic eDNA provided in an embodiment of the present invention; Figure 2 A block diagram of a biodiversity analysis system based on aquatic eDNA provided by an embodiment of the present invention; Figure 3 A diagram of a river structure in a water environment provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] like Figure 1 As shown, the present invention provides a method for analyzing biodiversity based on aquatic eDNA, comprising the following steps: Set up eDNA sampling points at the downstream of each river in the water environment and obtain eDNA data at each eDNA sampling point; Project the eDNA data onto the river distribution map of the water environment to obtain the eDNA spatial distribution map of the water environment; The eDNA data is associated with the flow characteristics of the river, and a rapid biodiversity analysis model is obtained for determining the eDNA data of the downstream river based on the eDNA data of the upstream river and the flow characteristics of the upstream river to the downstream river; The eDNA spatial distribution map is combined with the rapid biodiversity analysis model to obtain a rapid biodiversity analysis map.

[0020] The present invention obtains a spatial distribution map of eDNA data by spatially combining eDNA data with a river distribution map, and fully grasps the spatial distribution of eDNA data, so that the distribution of biodiversity in the water environment can be grasped according to the distribution of eDNA.

[0021] The present invention can analyze various eDNA circulation conditions in the river network, that is, the circulation conditions between the eDNA in the upper basin and the eDNA in the lower basin, in the eDNA spatial distribution map of the entire water environment. In order to understand the eDNA circulation conditions of the river network, the eDNA data is associated with the water flow characteristics of the river, and a rapid biodiversity analysis model is constructed to determine the eDNA data of the downstream river based on the eDNA data of the upstream river and the water flow characteristics from the upstream river to the downstream river.

[0022] The present invention measures the water flow correlation between the upper basin and the lower basin in the water environment through mutual information. If the correlation is high, the possibility of eDNA flowing from the upper basin to the lower basin is high, and if the correlation is low, the possibility of eDNA flowing from the upper basin to the lower basin is low.

[0023] The present invention encapsulates the possibility of eDNA circulation from the upper basin to the lower basin into the model, and obtains the mapping relationship between the eDNA data of the upstream river and the water flow characteristics from the upstream river to the downstream river. That is, on the premise of the possibility of eDNA circulation from the upper basin to the lower basin, the eDNA data of the downstream river is calculated through the eDNA data of the upstream river, so that only the eDNA data of the upstream river and the water flow that is easy to monitor can be detected. According to the rapid analysis model of biodiversity, the eDNA data of the downstream river is directly calculated and obtained, without the tedious detection process of the eDNA data of the downstream river, thereby realizing the rapid analysis and grasp of the eDNA data in the entire water environment.

[0024] The present invention constructs a rapid biodiversity analysis model, which can quickly grasp the eDNA data in the entire water environment by monitoring the eDNA data of the upstream river at any time, and ultimately achieve rapid analysis of the biodiversity of the water environment.

[0025] Furthermore, in order to improve the prediction accuracy of the rapid analysis model of biodiversity, the present invention only needs to obtain the eDNA data of the upstream river, and the eDNA data of all downstream rivers under the upstream river can be obtained through the eDNA data of the upstream river. The present invention sets the loss function of training the rapid analysis model of biodiversity to minimize the prediction loss between the upstream and downstream levels, thereby improving the prediction performance of the model. By monitoring the eDNA data of the upstream river, the eDNA data in the entire water environment can be quickly grasped.

[0026] Methods for obtaining eDNA data at eDNA sampling sites include: Collect surface water samples at the eDNA sampling points to obtain water environment samples at the eDNA sampling points; The water environment sample was filtered through a mesh sieve, and the solid matter was collected in a sterile centrifuge tube. Sodium chloride solution and benzalkonium chloride solution were added in a ratio of 1g:1mL:1mL to obtain a solid sample. The filtered water sample was then collected in a sterile ziplock bag and Add solid benzalkonium chloride to obtain a liquid sample; Add proteinase K solution, sodium dodecyl sulfate solution, and ethylenediaminetetraacetic acid solution to the solid sample and mix well. oscillation , centrifuge, retain the supernatant, then mix the supernatant with the liquid sample, filter with a filter membrane, and intercept the eDNA in the water sample onto the filter membrane; The eDNA on the filter membrane was extracted using an eDNA extraction kit, and the concentration and quality of the extracted genomic eDNA were tested. Then, a PE gene library was constructed, bridge PCR was performed, and finally, metagenomic sequencing was performed using the Illumina HiSeq high-throughput sequencing platform to obtain the eDNA data of the eDNA sampling point.

[0027] The method for constructing the spatial distribution map of eDNA in the water environment includes: In the river distribution map of the water environment, the eDNA data are displayed at the corresponding rivers according to the eDNA sampling points to obtain the eDNA spatial distribution map of the water environment.

[0028] The method for constructing a rapid biodiversity analysis model includes: The eDNA data of the upstream rivers in the eDNA spatial distribution map and the water flow from the upstream rivers to the downstream rivers are used as the input of the neural network, and the eDNA data of the downstream rivers in the eDNA spatial distribution map are used as the output of the neural network; Use neural networks to map inputs and outputs to train and obtain a rapid biodiversity analysis model; The rapid analysis model of biodiversity is: Downstream-eDNA=BP(Upstream-eDNA, Q-UptoDown); Where Downstream-eDNA is the predicted value of the eDNA data of the downstream river, Upstream-eDNA is the eDNA data of the upstream river, Q-UptoDown is the flow characteristics from the upstream river to the downstream river, and BP is a neural network.

[0029] In order to improve the prediction accuracy of the rapid analysis model of biodiversity, the present invention only needs to obtain the eDNA data of the upstream river, and the eDNA data of all downstream rivers under the upstream river can be obtained through the eDNA data of the upstream river. The loss function of training the rapid analysis model of biodiversity is set to minimize the prediction loss between the upstream and downstream levels, thereby improving the prediction performance of the model. By monitoring the eDNA data of the upstream river, the eDNA data in the entire water environment can be quickly mastered, as follows: The loss function for training the rapid biodiversity analysis model is: ; Where L is the loss function, (Downstream-eDNA)1 is the predicted value of the eDNA data of the downstream river in the first layer of the water environment, (Downstream-eDNA-real)1 is the true value of the eDNA data of the downstream river in the first layer of the water environment, and (Downstream-eDNA-real) k is the true value of the eDNA data of the downstream river in the kth layer of the water environment, (Downstream-eDNA) k-1 is the predicted value of the eDNA data of the downstream river in the k-1th layer of the water environment, (Q-UptoDown) k is the flow characteristics from the upstream river to the downstream river at the kth level in the water environment, and n is the total number of upstream and downstream levels of rivers in the water environment; BP((Downstream-eDNA) k-1 ,(Q-UptoDown) k ) is the predicted value of the eDNA data of the downstream river in the kth layer of the water environment; as well as are all L1 norm forms. Figure 3 As shown, the present invention detects the eDNA data of rivers h11-h52 in the water environment and the water flow sequence between all rivers between h11-h52, so as to obtain sample data for building a rapid analysis model of biodiversity. The data structure of the sample data is: (eDNA of the upstream river h11, eDNA of the downstream river h21, and water flow characteristics between h11-h21), (eDNA of the upstream river h11, eDNA of the downstream river h22, and water flow characteristics between h11-h22), (eDNA of the upstream river h21, eDNA of the downstream river h31, and water flow characteristics between h21-h31), (eDNA of the upstream river h32, eDNA of the downstream river h41, and water flow characteristics between h31-h41). To avoid redundancy, no exhaustive enumeration is given here.

[0030] The present invention constructs a rapid biodiversity analysis model based on sample data, so that the eDNA of the downstream river h21 can be obtained according to the eDNA of the upstream river h11 and the water flow characteristics between h11-h21, the eDNA of the downstream river h22 can be obtained according to the eDNA of the upstream river h11 and the water flow characteristics between h11-h22, and the eDNA of the downstream river h41 can be obtained according to the eDNA of the upstream river h32 and the water flow characteristics between h32-h41, so as to determine the eDNA data of the downstream river according to the eDNA data of the upstream river and the water flow characteristics from the upstream river to the downstream river.

[0031] Furthermore, the loss function set by the present invention can make the loss of the predicted value of the eDNA data of river h21 obtained through the eDNA data of river h11 the smallest, that is, the closest to the true value of the eDNA data of river h21, and the loss of the predicted value of the eDNA data of river h22 obtained through the eDNA data of river h11 the smallest, that is, the closest to the true value of the eDNA data of river h22, and reach the second upstream and downstream level (rivers h21 to h31 and h32, rivers h22 to h33 and h44) downstream. The eDNA data prediction values ​​of river h31 and river h32 obtained by the data prediction value have the smallest loss, and the eDNA data prediction values ​​of river h33 and river h34 obtained by the eDNA data prediction value of river h22 have the smallest loss, and go downstream to the third upstream and downstream level (river h32 to h41 and h42), and the eDNA data prediction values ​​of river h41 and river h42 obtained by the eDNA data prediction value of river h32 have the smallest loss, and so on for rivers h51 and h52, until all upstream and downstream levels in the water environment are reached.

[0032] The loss function constructed by the present invention is used to train the rapid analysis model of biodiversity, so that the predicted values ​​of the eDNA data of each river between h21-h52 that are closest to the true value can be gradually calculated based on the eDNA of river h11, thereby realizing that the rapid analysis model of biodiversity can grasp the eDNA data of all rivers in the entire water environment through the eDNA data of river h11, with fast calculation and simplified process.

[0033] Moreover, after constructing the rapid biodiversity analysis model, if you want to conduct biodiversity analysis of the entire water environment, you can do it at any time. You only need to detect the eDNA data of river h11, thus enhancing the flexibility of biodiversity analysis.

[0034] Methods for obtaining flow characteristics from upstream rivers to downstream rivers include: During the collection period of eDNA data, the velocity-area method (an existing mature method, which will not be described in detail here) is used to obtain the flow sequence of the upstream river and the flow sequence of the downstream river at the sampling points of the upstream river and the downstream river; Mutual information is used to quantify the degree of dependence of the flow sequence of the downstream river on the flow sequence of the upstream river, and the degree of dependence of the flow sequence of the downstream river on the flow sequence of the upstream river is taken as the flow characteristic from the upstream river to the downstream river; The quantitative formula for the flow characteristics from the upstream river to the downstream river is: ; Where Q-UptoDown is the flow characteristic from the upstream river to the downstream river, is the flow sequence of the upstream river, is the flow sequence of the downstream river, for and The joint probability distribution of for The marginal probability distribution of for The marginal probability distribution of .

[0035] The present invention measures the water flow correlation between the upper basin and the lower basin in the water environment through mutual information. If the correlation is high, the possibility of eDNA flowing from the upper basin to the lower basin is high, and if the correlation is low, the possibility of eDNA flowing from the upper basin to the lower basin is low.

[0036] The present invention encapsulates the possibility of eDNA circulation from the upper basin to the lower basin into the model, and obtains the mapping relationship between the eDNA data of the upstream river and the water flow characteristics from the upstream river to the downstream river. That is, on the premise of the possibility of eDNA circulation from the upper basin to the lower basin, the eDNA data of the downstream river is calculated through the eDNA data of the upstream river, so that only the eDNA data of the upstream river and the water flow that is easy to monitor can be detected. According to the rapid analysis model of biodiversity, the eDNA data of the downstream river is directly calculated and obtained, without the tedious detection process of the eDNA data of the downstream river, thereby realizing the rapid analysis and grasp of the eDNA data in the entire water environment.

[0037] The methods for constructing a rapid biodiversity analysis map include: In the eDNA spatial distribution map of the water environment, the rapid analysis model of biodiversity is encapsulated in the corresponding eDNA data to obtain the rapid analysis map of biodiversity.

[0038] The biodiversity rapid analysis map can visualize and operationalize rapid analysis, improve the simplicity of analysis, have low professional requirements for operators, and has strong versatility.

[0039] like Figure 2 As shown, the present invention provides a biodiversity analysis system based on aquatic eDNA, which is applied to a biodiversity analysis method based on aquatic eDNA. The system includes: The sampling unit is used to set up eDNA sampling points at the downstream of each river in the water environment and obtain eDNA data of each eDNA sampling point; A data processing unit, used to associate the eDNA data with the flow characteristics of the river, and obtain a biodiversity rapid analysis model for determining the eDNA data of the downstream river based on the eDNA data of the upstream river and the flow characteristics of the upstream river to the downstream river; A graphics construction unit is used to project the eDNA data onto a river distribution map of the water environment to obtain an eDNA spatial distribution map of the water environment; And, it is used to combine the eDNA spatial distribution map with the rapid analysis model of biodiversity to obtain a rapid analysis map of biodiversity.

[0040] The biodiversity rapid analysis model is encapsulated into the biodiversity rapid analysis map through a selection button control, and the biodiversity rapid analysis model is activated to obtain eDNA data by selecting the selection button control.

[0041] The eDNA data is displayed in the biodiversity rapid analysis map through a digital display frame control, and the digital display frame control is interconnected with the output end of the biodiversity rapid analysis model.

[0042] The method for constructing a rapid analysis model of biodiversity by the data processing unit includes: The eDNA data of the upstream rivers in the eDNA spatial distribution map and the water flow from the upstream rivers to the downstream rivers are used as the input of the neural network, and the eDNA data of the downstream rivers in the eDNA spatial distribution map are used as the output of the neural network; Use neural networks to map inputs and outputs to train and obtain a rapid biodiversity analysis model; The rapid analysis model of biodiversity is: Downstream-eDNA=BP(Upstream-eDNA, Q-UptoDown); In the formula, Downstream-eDNA is the predicted value of the eDNA data of the downstream river, Upstream-eDNA is the eDNA data of the upstream river, Q-UptoDown is the flow characteristics from the upstream river to the downstream river, and BP is a neural network; Among them, the loss function for training the rapid analysis model of biodiversity is: ; Where L is the loss function, (Downstream-eDNA)1 is the predicted value of the eDNA data of the downstream river in the first layer of the water environment, (Downstream-eDNA-real)1 is the true value of the eDNA data of the downstream river in the first layer of the water environment, and (Downstream-eDNA-real) k is the true value of the eDNA data of the downstream river in the kth layer of the water environment, (Downstream-eDNA) k-1 is the predicted value of the eDNA data of the downstream river in the k-1th layer of the water environment, (Q-UptoDown) k is the flow characteristics from the upstream river to the downstream river at the kth level in the water environment, and n is the total number of upstream and downstream levels of rivers in the water environment; BP((Downstream-eDNA) k-1 ,(Q-UptoDown) k ) is the predicted value of the eDNA data of the downstream river in the kth layer of the water environment; as well as All are L1 norm forms.

[0043] The present invention projects eDNA data onto a river distribution map of a water environment to obtain an eDNA spatial distribution map of the water environment, which can grasp the eDNA distribution in the entire water environment. Then, by combining the eDNA spatial distribution map with a rapid biodiversity analysis model, a rapid biodiversity analysis map is obtained, which can realize rapid analysis of eDNA distribution and ultimately realize rapid analysis of biodiversity in the entire water environment.

[0044] The neural network for constructing the rapid analysis model of biodiversity can use the BP (back propagation) neural network. The BP (back propagation) neural network is a concept proposed by scientists led by Rumelhart and McClelland. It is a multi-layer feedforward neural network trained according to the error back propagation algorithm and is one of the most widely used neural network models.

[0045] BP neural network has the ability to classify patterns of any complexity and excellent multi-dimensional function mapping ability. It can learn certain rules through its own training and obtain the result closest to the expected output value when the input value is given, that is, determine the mapping relationship between input and output. The BP algorithm includes two processes: forward propagation of signals and reverse propagation of errors. That is, when calculating the error output, it is done from input to output, while adjusting the weights and thresholds from output to input. During forward propagation, the input signal acts on the output node through the hidden layer, and after nonlinear transformation, the output signal is generated. If the actual output does not match the expected output, the error reverse propagation process is entered. Error back propagation is to propagate the output error through the hidden layer to the input layer layer by layer, and distribute the error to all units in each layer, and use the error signal obtained from each layer as the basis for adjusting the weights of each unit. By adjusting the connection strength between the input node and the hidden layer node, the connection strength between the hidden layer node and the output node, and the threshold, the error is reduced along the gradient direction. After repeated learning and training, the network parameters (weights and thresholds) corresponding to the minimum error are determined, and the training stops. At this point, the trained neural network can process the input information of similar samples, and obtain the output information with the smallest output error and after nonlinear transformation.

[0046] Utilizing the function mapping properties of the BP neural network, the present invention applies the BP neural network to the process of constructing a rapid analysis model for biodiversity, and trains the BP neural network to learn the mapping relationship between the eDNA data of the upstream river and the flow characteristics from the upstream river to the downstream river to determine the eDNA data of the downstream river.

[0047] The above embodiments are only exemplary embodiments of the present application and are not intended to limit the present application. The protection scope of the present application is defined by the claims. Those skilled in the art may make various modifications or equivalent substitutions to the present application within the essence and protection scope of the present application, and such modifications or equivalent substitutions shall also be deemed to fall within the protection scope of the present application.

Claims

1. A biodiversity analysis method based on aquatic eDNA, characterized by: The following steps are involved: Set up eDNA sampling points at the downstream of each river in the water environment and obtain eDNA data at each eDNA sampling point; Project the eDNA data onto the river distribution map of the water environment to obtain the eDNA spatial distribution map of the water environment; The eDNA data is associated with the flow characteristics of the river, and a rapid biodiversity analysis model is obtained for determining the eDNA data of the downstream river based on the eDNA data of the upstream river and the flow characteristics of the upstream river to the downstream river; The eDNA spatial distribution map is combined with the rapid biodiversity analysis model to obtain a rapid biodiversity analysis map.

2. The biodiversity analysis method based on aquatic eDNA according to claim 1, characterized in that: Methods for obtaining eDNA data at eDNA sampling sites include: Collect surface water samples at the eDNA sampling points to obtain water environment samples at the eDNA sampling points; The water environment sample was filtered through a mesh sieve, the solid matter was collected with a sterile centrifuge tube, and sodium chloride solution and benzalkonium chloride solution were added at a ratio of 1 g:1 mL:1 mL to obtain a solid sample. The filtered water sample was then collected in a sterile ziplock bag and emptied according to the ratio. Add solid benzalkonium chloride to obtain a liquid sample; Add proteinase K solution, sodium dodecyl sulfate solution and ethylenediaminetetraacetic acid solution to the solid sample and mix well. oscillation , centrifuging, retaining the supernatant, then mixing the supernatant with the liquid sample, filtering with a filter membrane, and intercepting the eDNA in the water sample onto the filter membrane; The eDNA on the filter membrane was extracted using an eDNA extraction kit, and the concentration and quality of the extracted genomic eDNA were tested. Then, a PE gene library was constructed, bridge PCR was performed, and finally, metagenomic sequencing was performed using the Illumina HiSeq high-throughput sequencing platform to obtain the eDNA data of the eDNA sampling point.

3. The biodiversity analysis method based on aquatic eDNA according to claim 1, characterized in that: The method for constructing the eDNA spatial distribution map of the water environment comprises: In the river distribution map of the water environment, the eDNA data is displayed at the corresponding river according to the eDNA sampling points to obtain the eDNA spatial distribution map of the water environment.

4. The biodiversity analysis method based on aquatic eDNA according to claim 3, characterized in that: The method for constructing the biodiversity rapid analysis model comprises: The eDNA data of the upstream rivers in the eDNA spatial distribution map and the water flow from the upstream rivers to the downstream rivers are used as the input of the neural network, and the eDNA data of the downstream rivers in the eDNA spatial distribution map are used as the output of the neural network; Using a neural network to map input and output to train, to obtain the biodiversity rapid analysis model; The biodiversity rapid analysis model is: Downstream-eDNA=BP(Upstream-eDNA, Q-UptoDown); In the formula, Downstream-eDNA is the predicted value of the eDNA data of the downstream river, Upstream-eDNA is the eDNA data of the upstream river, Q-UptoDown is the flow characteristics from the upstream river to the downstream river, and BP is a neural network; Among them, the loss function for training the rapid analysis model of biodiversity is: ； Where L is the loss function, (Downstream-eDNA)1 is the predicted value of the eDNA data of the downstream river in the first layer of the water environment, (Downstream-eDNA-real)1 is the true value of the eDNA data of the downstream river in the first layer of the water environment, and (Downstream-eDNA-real) k is the true value of the eDNA data of the downstream river in the kth layer of the water environment, (Downstream-eDNA) k-1 is the predicted value of the eDNA data of the downstream river in the k-1th layer of the water environment, (Q-UptoDown) k is the flow characteristics from the upstream river to the downstream river at the kth level in the water environment, and n is the total number of upstream and downstream levels of rivers in the water environment; BP((Downstream-eDNA) k-1 ,(Q-UptoDown) k ) is the predicted value of the eDNA data of the downstream river in the kth layer of the water environment; as well as All are L1 norm forms.

5. The biodiversity analysis method based on aquatic eDNA according to claim 4, characterized in that: Methods for obtaining flow characteristics from upstream rivers to downstream rivers include: During the collection period of eDNA data, the velocity-area method was used to obtain the flow rate series of the upstream river and the flow rate series of the downstream river at the sampling points of the upstream river and the downstream river; The mutual information is used to quantify the degree of dependence of the water flow sequence of the downstream river on the water flow sequence of the upstream river, and the degree of dependence of the water flow sequence of the downstream river on the water flow sequence of the upstream river is used as the water flow characteristic from the upstream river to the downstream river; The quantitative formula for the flow characteristics from the upstream river to the downstream river is: ； Where Q-UptoDown is the flow characteristic from the upstream river to the downstream river, is the flow sequence of the upstream river, is the flow sequence of the downstream river, for and The joint probability distribution of for The marginal probability distribution of for The marginal probability distribution of .

6. The biodiversity analysis method based on aquatic eDNA according to claim 5, characterized in that: The method for constructing the biodiversity rapid analysis map comprises: In the eDNA spatial distribution map of the water environment, the biodiversity rapid analysis model is encapsulated in the corresponding eDNA data to obtain the biodiversity rapid analysis map.

7. A biodiversity analysis system based on aquatic eDNA, characterized in that: A method for analyzing biodiversity based on aquatic eDNA as described in any one of claims 1 to 6, the system comprising: The sampling unit is used to set up eDNA sampling points at the downstream of each river in the water environment and obtain eDNA data of each eDNA sampling point; A data processing unit, used to associate the eDNA data with the flow characteristics of the river, and obtain a biodiversity rapid analysis model for determining the eDNA data of the downstream river based on the eDNA data of the upstream river and the flow characteristics of the upstream river to the downstream river; A graphics construction unit is used to project the eDNA data onto a river distribution map of the water environment to obtain an eDNA spatial distribution map of the water environment; And, it is used to combine the eDNA spatial distribution map with the rapid analysis model of biodiversity to obtain a rapid analysis map of biodiversity.

8. The biodiversity analysis system based on aquatic eDNA according to claim 7, characterized in that: The biodiversity rapid analysis model is encapsulated into the biodiversity rapid analysis map through a selection button control, and the biodiversity rapid analysis model is activated by selecting the selection button control to obtain eDNA data.

9. The biodiversity analysis system based on aquatic eDNA according to claim 7, characterized in that: The eDNA data is displayed in a biodiversity rapid analysis map through a digital display frame control, and the digital display frame control is interconnected with an output end of a biodiversity rapid analysis model.

10. The biodiversity analysis system based on aquatic eDNA according to claim 7, characterized in that: The method for constructing a rapid analysis model of biodiversity by the data processing unit includes: The eDNA data of the upstream rivers in the eDNA spatial distribution map and the water flow from the upstream rivers to the downstream rivers are used as the input of the neural network, and the eDNA data of the downstream rivers in the eDNA spatial distribution map are used as the output of the neural network; Using a neural network to map input and output to train, to obtain the biodiversity rapid analysis model; The biodiversity rapid analysis model is: Downstream-eDNA=BP(Upstream-eDNA, Q-UptoDown); In the formula, Downstream-eDNA is the predicted value of the eDNA data of the downstream river, Upstream-eDNA is the eDNA data of the upstream river, Q-UptoDown is the flow characteristics from the upstream river to the downstream river, and BP is a neural network; Among them, the loss function for training the rapid analysis model of biodiversity is: ； Where L is the loss function, (Downstream-eDNA)1 is the predicted value of the eDNA data of the downstream river in the first layer of the water environment, (Downstream-eDNA-real)1 is the true value of the eDNA data of the downstream river in the first layer of the water environment, and (Downstream-eDNA-real) k is the true value of the eDNA data of the downstream river in the kth layer of the water environment, (Downstream-eDNA) k-1 is the predicted value of the eDNA data of the downstream river in the k-1th layer of the water environment, (Q-UptoDown) k is the flow characteristics from the upstream river to the downstream river at the kth level in the water environment, and n is the total number of upstream and downstream levels of rivers in the water environment; BP((Downstream-eDNA) k-1 ,(Q-UptoDown) k ) is the predicted value of the eDNA data of the downstream river in the kth layer of the water environment; as well as All are L1 norm forms.

Images