CN114283882B - Non-destructive poultry egg quality character prediction method and system - Google Patents
Non-destructive poultry egg quality character prediction method and system Download PDFInfo
- Publication number
- CN114283882B CN114283882B CN202111677548.6A CN202111677548A CN114283882B CN 114283882 B CN114283882 B CN 114283882B CN 202111677548 A CN202111677548 A CN 202111677548A CN 114283882 B CN114283882 B CN 114283882B
- Authority
- CN
- China
- Prior art keywords
- egg
- dna sequence
- data set
- poultry
- coding
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a non-destructive poultry egg quality character prediction method and a non-destructive poultry egg quality character prediction system, wherein the method comprises the steps of collecting an image data set and a genotype data set of a sample poultry egg; training a preset egg quality character prediction model according to the image data set and the genotype data set to obtain a trained egg quality character prediction model; and predicting the phenotypic character of the target egg from the target image and the target DNA sequence of the target egg according to the trained egg quality character prediction model. According to the method, the external image data of the poultry egg and the genotype data of the corresponding laying hen are combined, the phenotypic characters of the poultry egg which can be measured only after being destroyed can be determined through the lossless inversion of the egg quality character prediction model, and the accuracy of the result of the phenotypic character prediction is improved. Compared with the existing scheme, the method is more accurate, simple, convenient and feasible.
Description
Technical Field
The invention relates to the technical field of nondestructive testing of egg quality, in particular to a nondestructive egg quality character prediction method and a nondestructive egg quality character prediction system.
Background
Egg quality traits are important economic traits of poultry concern in breeding and production, and the egg quality phenotypic traits of poultry include both extrinsic and intrinsic qualities, are typically quantitative traits controlled by a micro-effective polygene, and are influenced by multiple factors such as environment, nutrition and feed management.
Although some detection devices and laboratory methods can detect corresponding indexes of eggs (eggs produced by poultry) at home and abroad at present, the determination of the quality of the eggs is destructive, the existing chemical methods for detecting the quality of the eggs break the eggs to determine some physicochemical factors such as the quality of egg white, the quality of egg yolk and the like, which is undoubtedly not beneficial to the actual breeding production, and the damaged eggs cannot be hatched, so that the selection or influence of the quality characters of the eggs on the growth performance or the reproductive performance of breeding individuals is difficult to effectively reveal.
Disclosure of Invention
The present invention is directed to at least solving the problems of the prior art. Therefore, the invention provides a nondestructive poultry egg quality character prediction method and a nondestructive poultry egg quality character prediction system. The phenotypic characters of the poultry eggs can be determined only after the poultry eggs are damaged in a lossless reverse manner, and the accuracy of the phenotypic character prediction result is improved.
The invention provides a nondestructive poultry egg quality character prediction method, which comprises the following steps:
acquiring an image data set of sample poultry eggs and a genotype data set of corresponding laying hens, wherein the image data set comprises a plurality of poultry egg images marked with egg quality phenotypic characters, the genotype data set comprises DNA sequences of the laying hens, and the DNA sequences are obtained by whole genome DNA (deoxyribonucleic acid) re-sequencing or targeted sequencing;
training a preset egg quality character prediction model according to the image data set and the genotype data set to obtain a trained egg quality character prediction model;
and predicting the phenotypic character of the target egg from the target image and the target DNA sequence of the target egg according to the trained egg quality character prediction model.
According to the embodiment of the invention, at least the following technical effects are achieved:
the method overcomes the detection limitation of the existing detection device and the complexity, destructiveness and other problems of a laboratory method of the existing detection device, utilizes an artificial intelligence technology, combines the external image data and the genotype data of the poultry egg, utilizes the external characteristics and the genotype information of the fused external image data, and improves the accuracy of the result of phenotypic character prediction by constructing the phenotypic characters of the poultry egg which can be measured only after being destroyed through the lossless inversion of an egg quality character prediction model. Compared with the existing scheme, the prediction method is more accurate, simple, convenient and feasible.
According to some embodiments of the invention, the image of the avian egg comprises a visible light image, an infrared image and a hyperspectral image, the hyperspectral image comprising spectral images of a plurality of wave bands.
According to some embodiments of the invention, the training of a preset egg quality trait prediction model from the image dataset and the genotype dataset comprises:
extracting image features from the image dataset according to a deep convolutional neural network;
coding the DNA sequence in the genotype dataset according to a DNA coding algorithm to obtain a coding result, and extracting DNA sequence characteristics from the coding result according to a deep recurrent neural network;
and fusing the image characteristics and the DNA sequence characteristics to obtain fusion characteristics, and predicting the phenotypic characters of the sample poultry eggs from the fusion characteristics according to a full-link layer and a softmax layer.
According to some embodiments of the invention, the deep convolutional neural network comprises, in order: convolutional layers, pooling layers, regularization layers, recursive layers, and fully-connected layers.
According to some embodiments of the present invention, the encoding the DNA sequence in the genotype dataset according to a DNA coding algorithm to obtain a coding result, and extracting DNA sequence features from the coding result according to a deep recurrent neural network, includes:
coding the DNA sequence according to Onehot to obtain a first coding result, coding the DNA sequence according to DBE to obtain a second coding result, coding the DNA sequence according to NCP to obtain a third coding result, and coding the DNA sequence according to k _ mer to obtain a fourth coding result;
performing sequence ensemble learning on the first coding result to the fourth coding result to obtain a sequence ensemble learning result;
and inputting the sequence integrated learning result into the deep recurrent neural network to obtain the DNA sequence characteristics extracted by the deep recurrent neural network. According to some embodiments of the invention, sequence ensemble learning is performed on the first to fourth encoding results according to XGBoost.
According to some embodiments of the invention, the phenotypic trait includes protein content, protein pH, harderian units, lecithin content, egg yolk color, and egg yolk proportion.
In a second aspect of the present invention, there is provided a nondestructive poultry egg quality trait prediction system, comprising:
the system comprises a sample acquisition unit, a sample analysis unit and a control unit, wherein the sample acquisition unit is used for acquiring an image data set of a sample egg and a genotype data set of a corresponding laying hen of the sample egg, the image data set comprises a plurality of egg images marked with egg quality phenotypic characters, the genotype data set comprises a DNA sequence of the laying hen, and the DNA sequence is obtained by whole genome DNA re-sequencing or targeted sequencing;
the model training module is used for training a preset egg quality character prediction model according to the image data set and the genotype data set to obtain a trained egg quality character prediction model;
and the model prediction unit is used for predicting the phenotypic character of the target egg from the target image and the target DNA sequence of the target egg according to the trained egg quality character prediction model.
In a third aspect of the invention, an electronic device is provided, comprising at least one control processor and a memory for communicative connection with the at least one control processor; the memory stores instructions executable by the at least one control processor to enable the at least one control processor to perform the non-destructive egg quality trait prediction method described above.
In a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer-executable instructions for causing a computer to perform the non-destructive method for predicting quality traits of poultry eggs described above.
It should be noted that the advantageous effects between the second to fourth aspects of the present invention and the prior art are the same as the advantageous effects between the nondestructive poultry egg quality trait prediction method and the prior art, and will not be described in detail here.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a schematic flow chart of a non-destructive method for predicting quality and properties of poultry eggs according to an embodiment of the present invention;
fig. 2 is a general flow chart of a method for predicting a quality trait of a non-destructive avian egg according to an embodiment of the present invention;
fig. 3 is a general flow chart diagram of a method for predicting a non-destructive egg quality trait according to another embodiment of the present invention;
fig. 4 is a schematic structural diagram of a nondestructive egg quality trait prediction system according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of a readable storage medium according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
In the description of the present invention, it is to be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "axial", "radial", "circumferential", and the like, indicate orientations and positional relationships based on the orientations and positional relationships shown in the drawings, and are used merely for convenience in describing the present invention and for simplicity in description, and do not indicate or imply that the device or element so referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore, should not be construed as limiting the present invention.
Furthermore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present invention, "a plurality" means two or more unless otherwise specified.
The egg quality phenotype character is an important economic character concerned in breeding and production of poultry, and the egg quality character of poultry (laying hens) comprises an external quality character and an internal quality character, is a typical quantitative character controlled by a micro-effective polygene, and is influenced by a plurality of factors such as environment, nutrition, feeding management and the like. Although some detection devices and laboratory methods can detect corresponding indexes of poultry eggs (eggs produced by laying hens) at home and abroad at present, the determination of the quality of the poultry eggs is destructive, and the existing chemical methods for detecting the quality of the poultry eggs break the eggs to determine some physicochemical factors (such as protein quality (protein content, protein pH value and Haugh unit), yolk quality (lecithin content, yolk color and yolk proportion) or other indexes (chemical components, functional characteristics of the eggs, blood spots, meat spots, taste, smell, microbial conditions and the like)), which is undoubtedly not beneficial to actual breeding production, and the destroyed eggs cannot be hatched any more, and the selection or influence of the properties of the eggs on the growth performance or reproduction performance of breeding individuals is difficult to effectively reveal. At present, most of nondestructive detection methods for the quality and the properties of eggs depend on appearances or some penetrating light and spectrum methods, and the inherent chemical quality of the eggs is difficult to accurately predict.
The invention aims to overcome the detection limitation of the existing detection device and the problems of complexity, destructiveness and the like of a laboratory method. By using an artificial intelligence technology, combining external image data of the poultry eggs with genotype data of the laying hens and fusing external characteristics and genotype data of the external image data, the phenotype characters of the poultry eggs which can be measured only after being destroyed can be subjected to lossless inversion, and the accuracy of the result of phenotype character prediction is improved.
Referring to fig. 1 to 3, an embodiment of the present invention provides a method for predicting a quality trait of a non-destructive avian egg, including the following steps:
s200, collecting an image data set of a sample egg and a genotype data set of a corresponding layer of the sample egg, wherein the image data set comprises a plurality of egg images marked with egg quality phenotype characters, the genotype data set comprises a DNA sequence of the layer, and the DNA sequence is obtained by whole genome DNA re-sequencing or targeted sequencing.
The genome DNA sequence is composed of single nucleotide (A, T, C or G), is also the basic unit of inheritance, determines the character characteristics of organisms, and therefore, the selection of favorable DNA sequence or site is the main means of animal and plant molecular breeding. In the embodiment of the application, the phenotypic traits of egg quality include, but are not limited to, egg white quality (protein content, egg white pH value, Ha's unit) and yolk quality (lecithin content, yolk color, yolk proportion), and the like, therefore, a candidate gene marker related to the egg quality trait is firstly determined (a related candidate gene marker is selected according to the phenotypic trait to be predicted), and a DNA sequence of the candidate gene marker is obtained. It should be noted that the relevant data of the DNA sequence comes from a breeding center, and the breeding center collects the data of the corresponding genotype and phenotypic trait in the actual breeding production to guide the breeding work.
In some embodiments, the image of the avian egg comprises a visible light image, an infrared image, and a hyperspectral image, the hyperspectral image comprising a plurality of bands of spectral images. In a related research scheme for phenotypic characters of egg quality, a visible light image (RGB image) and an infrared image are generally used for research, i.e., the visible light image and the infrared image are used as input data of a neural network, and although more features can be extracted from the visible light image and the infrared image, many available features exist in a complex situation, and the available features are not fully utilized. The hyperspectral images have richer spectral information (information with more dimensions) in all bands, so that the hyperspectral images can reflect the slight changes of the physical characteristics of different poultry eggs, can detect the changes of the internal structures and chemical components of the poultry eggs, and can extract richer features from the hyperspectral images through the models, thereby improving the identification precision of the models. It should be noted that the image data may originate from a breeding center and may be acquired by an imaging device.
And S400, training a preset egg quality character prediction model according to the image data set and the genotype data set to obtain the trained egg quality character prediction model. Step S400 specifically includes the following steps:
and S410, extracting image features from the image data set according to the deep convolutional neural network. In some embodiments, the present embodiment sets the deep convolutional neural network to sequentially include: convolution layer → pooling layer → regular layer (dropout layer or bn layer) → recursive layer → regular layer → fully connected layer (dimension reduction). It is worth mentioning that the extraction of image features based on the deep convolution neural network is common knowledge in the field of neural network technology, and will not be described in detail here.
And step S420, coding the DNA sequence in the genotype dataset according to a DNA coding algorithm to obtain a coding result, and extracting DNA sequence characteristics from the coding result according to a deep recurrent neural network. Specifically, step S420 includes the steps of:
step S4201, coding the DNA sequence according to Onehot to obtain a first coding result, coding the DNA sequence according to DBE to obtain a second coding result, coding the DNA sequence according to NCP to obtain a third coding result, and coding the DNA sequence according to k _ mer to obtain a fourth coding result.
(1) Onehot is one of the important ways to encode classification tasks. Onehot encoding forms vector dimensions equal to the size of the dictionary. The coding method is that after all words in the dictionary are sequenced, the corresponding positions are coded as 1, and the rest positions are 0. For DNA sequences, there are a total of four base types: A. c, G, T, i.e. dictionary size is 4. The results of the encoding of these four bases: a = [0,0,0,1], C = [0,0,1,0], G = [0,1,0,0], T = [1,0,0,0 ].
(2) DBE encodes two bases adjacent to each other in a character string represented by a DNA sequence. As shown in Table 1 below, a vector of length 4 is used for each set of dinucleotides.
| Base | Encoding | Base | Encoding |
| AA | [0，0，0，0] | AT | [0，0，0，1] |
| AC | [0，0，1，0] | AG | [0，0，1，1] |
| TA | [0，1，0，0] | TT | [0，1，0，1] |
| TC | [0，1，1，0] | TG | [0，1，1，1] |
| CA | [1，0，0，0] | CT | [1，0，0，1] |
| CC | [1，0，1，0] | CG | [1，0，1，1] |
| GA | [1，1，0，0] | GT | [1，1，0，1] |
| GC | [1，1，1，0] | GG | [1，1，1，1] |
TABLE 1
(3) NCPs are encoded by three chemical properties of nucleotides in a DNA sequence. These three chemical properties are: ring structure type (purine or pyrimidine), functional group type (amino or keto), number of hydrogen bonds formed by base pairing (3 or 2). Coding of purine, amino, 2 hydrogen bonds to 1, pyrimidine, keto, 3 hydrogen bonds to 0, one can obtain the coding result for each base: a = [1,1,1], C = [0,1,0], G = [1,0,0], and T = [0,0,1 ].
(4) A k mer is a contiguous subsequence of length k in a string of characters represented by a DNA sequence. In the method example, frequency statistics is carried out on 2_ mer, 3_ mer and 4_ mer in the DNA sequence. There are 16 (AA, AC, AG, AT, CA, CC, CG, CT, GA, GC, GG, GT, TA, TC, TG, TT) for a total of 2 mers, 64 (AAA, AAC,.., GTT, TTT) for _mers, and 256 (AAAA, AAAC,.., GTTT, TTTT) for 4 mers. The 2_ mer, the 3_ mer and the 4_ mer are arranged in the order, the arranged frequency is used as a coding result, and the obtained characteristic is a one-dimensional vector with the length of 336.
In step S4201, four gene coding methods, i.e., Onehot, DBE, NCP, and k-mer, are selected to comprehensively characterize the DNA sequence, and it should be noted that the coding is to code the DNA sequence in a vector form, so as to be used as input of the neural network.
And step S4202, performing ensemble learning on the first coding result to the fourth coding result according to AdaBoost or XGBoost to obtain ensemble learning results.
In order to further improve the prediction performance, a sequence ensemble learning method is used for feature selection of coding results obtained by different coding modes to achieve the effect of reducing variance and deviation or improving prediction. For example, the XGBoost evaluates the importance of all the encoding results (the encoding result is a one-dimensional vector or a two-dimensional vector), increases the proportion of the encoding result with high importance, and reduces the proportion of the encoding result with low importance, for example, when the importance of the one-dimensional vector feature obtained by the k _ mer encoding is low, the XGBoost can be deleted.
And step S4203, extracting DNA sequence features from the ensemble learning result according to the deep recurrent neural network.
In steps S4201 and S4202, first, the DNA sequence is encoded into a vector form as an input vector of the neural network, then, the XGBoost is used to perform ensemble learning on all encoding results to improve the prediction performance, the features after ensemble learning are directly input into the deep recurrent neural network, and a fully connected layer (Dense/FC) in the deep recurrent neural network is provided to extract feature vectors similar to image features.
It should be noted that the embodiment of the present application does not make any improvement on the network structure of the deep recurrent neural network itself, and in a preferred embodiment, a standard deep recurrent neural network architecture is adopted, which has a plurality of inputs and a plurality of outputs (related to the number of features), each input and output corresponds to a standard recurrent neural network unit, each unit has two hidden layers, and each hidden layer has 100 nodes.
And S430, fusing the image characteristics and the DNA sequence characteristics to obtain fusion characteristics, and predicting the phenotypic characters of the sample poultry eggs from the fusion characteristics according to the full-link layer and the softmax layer. There are various fusion methods, and splicing (concat) is selected in this embodiment.
In a pre-set egg quality character prediction model, a deep Convolutional Neural Network (CNN) employs a weight sharing strategy to capture local patterns in egg images (visible light images, infrared images, and hyperspectral images of eggs), and a deep recurrent neural network (RNN or LSTM) can use its internal states (memory) to learn DNA sequence patterns. That is, in the designed model, a deep convolutional neural network is used to learn primitive features, and a deep recursive neural network is used to learn long-term dependencies between primitive features. After the model is built, the training data set (the image data set and the genotype data set in step S200) is used for training, and the training and optimization of the deep learning model are completed.
And S600, predicting the phenotypic character of the target egg from the target image and the target DNA sequence of the target egg according to the trained egg quality character prediction model.
After model training is completed, the trained egg quality character prediction model is used for predicting the phenotype character of the target egg from the target image (including visible light, infrared and hyperspectral images) and the target DNA sequence of the target egg. The phenotypic traits predicted here include the above mentioned egg white qualities (protein content, egg white pH, hardy unit) and egg yolk qualities (lecithin content, egg yolk colour, egg yolk proportion).
In related schemes, most of nondestructive detection methods for the quality and characters of the poultry eggs depend on appearance or some penetrating light and spectrum methods, and the internal quality of the poultry eggs is difficult to accurately predict; moreover, the existing chemical methods for detecting the quality of the poultry eggs break the eggs to detect some physicochemical factors, which are not beneficial to practical breeding production, and the damaged eggs cannot be hatched, so that the selection or influence of the quality characters of the eggs on the growth performance or reproductive performance of breeding individuals is difficult to effectively reveal. The method overcomes the detection limitation of the existing detection device and the complexity, destructiveness and other problems of the laboratory method, utilizes the artificial intelligence technology, combines the external image data of the eggs and the genotype data of the laying hens, and improves the accuracy of the result of the phenotype character prediction by constructing an egg quality character prediction model to perform lossless inversion to determine the phenotype characters of the eggs which can be detected only after being destroyed. Compared with the existing scheme, the prediction method is more accurate, simple, convenient and feasible.
Referring to fig. 4, an embodiment of the present invention provides a nondestructive egg quality trait prediction system, including a sample acquisition unit 1000, a model training module 2000, and a model prediction unit 3000:
the sample acquisition unit 1000 is configured to acquire an image dataset of a sample poultry egg and a genotype dataset of a layer corresponding to the sample poultry egg, the image dataset including a plurality of poultry egg images with marked egg quality phenotype traits, the genotype dataset including a DNA sequence of the layer, the DNA sequence being obtained by whole genome DNA re-sequencing or targeted sequencing;
the model training module 2000 is configured to train a preset egg quality condition prediction model according to the image data set and the genotype data set, so as to obtain a trained egg quality condition prediction model.
The model prediction unit 3000 is configured to predict a phenotypic trait of the target egg from the target image and the target DNA sequence of the target egg according to the trained egg quality trait prediction model.
In related schemes, most of nondestructive detection methods for the quality and characters of the poultry eggs depend on appearance or some penetrating light and spectrum methods, and the internal quality of the poultry eggs is difficult to accurately predict; moreover, the existing chemical methods for detecting the quality of the poultry eggs break the eggs to detect some physicochemical factors, which are not beneficial to practical breeding production, and the broken eggs cannot be hatched, so that the selection or influence of the quality characters of the eggs on the growth performance or reproductive performance of breeding individuals is difficult to effectively reveal. The system overcomes the detection limitation of the existing detection device and the problems of complexity, destructiveness and the like of a laboratory method of the existing detection device, utilizes an artificial intelligence technology, combines the external image data of eggs with the internal genotype data of the laying hens with physiological and biochemical characteristics such as egg quality genetic characteristics, biological regulation and the like, and improves the accuracy of the result of phenotypic character prediction by constructing an egg quality character prediction model to perform lossless inversion to determine the phenotypic characters of the eggs which can be measured only after being destroyed. Compared with the existing scheme, the prediction system is more accurate, simple, convenient and feasible. It should be noted that the embodiment of the present system and the embodiment of the method are based on the same inventive concept, and therefore, the related contents of the embodiment of the method are also applicable to the embodiment of the present system, and detailed descriptions are not repeated.
Referring to fig. 5, the present application further provides a computer device 301 comprising: a memory 310, a processor 320 and a computer program 311 stored on the memory 310 and executable on the processor, the processor 320 when executing the computer program 311 effecting: such as the above-mentioned nondestructive method for predicting the quality traits of poultry eggs.
The processor 320 and memory 310 may be connected by a bus or other means.
The memory 310, which is a non-transitory computer-readable storage medium, may be used to store non-transitory software programs as well as non-transitory computer-executable programs. Further, the memory 310 may include high speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, memory 310 may optionally include memory located remotely from the processor, which may be connected to the processor via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The non-transitory software programs and instructions required to implement the non-destructive egg quality trait prediction method of the above-described embodiments are stored in a memory, and when executed by a processor, perform the non-destructive egg quality trait prediction method of the above-described embodiments, e.g., performing the method steps S200-S600 in fig. 1 described above.
Referring to fig. 6, the present application further provides a computer-readable storage medium 401 storing computer-executable instructions 410, the computer-executable instructions 410 being configured to perform: such as the above-mentioned nondestructive method for predicting the quality traits of poultry eggs.
The computer-readable storage medium 401 stores computer-executable instructions 410, and the execution of the computer-executable instructions 410 by a processor or controller, for example, by a processor in the above-mentioned embodiment of the electronic device, causes the processor to perform the non-destructive egg quality attribute prediction method in the above-mentioned embodiment, for example, the method steps S200 to S600 in fig. 1 described above.
It will be understood by those of ordinary skill in the art that all or some of the steps, systems, and methods disclosed above may be implemented as software, firmware, hardware, or suitable combinations thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of data such as computer readable instructions, data structures, program modules or other data, as is well known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired data and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any data delivery media as known to one of ordinary skill in the art.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an illustrative embodiment," "an example," "a specific example," or "some examples" or the like mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
Claims (10)
1. A non-destructive poultry egg quality character prediction method is characterized by comprising the following steps:
acquiring an image data set of a sample poultry egg and a genotype data set of a corresponding laying hen of the sample poultry egg, wherein the image data set comprises a plurality of poultry egg images marked with egg quality phenotypic characters, the genotype data set comprises a DNA sequence of the laying hen, and the DNA sequence is obtained by whole genome DNA re-sequencing or targeted sequencing;
training a preset egg quality character prediction model according to the image data set and the genotype data set to obtain a trained egg quality character prediction model;
and predicting the phenotypic character of the target poultry egg from the target image and the target DNA sequence of the target poultry egg according to the trained egg quality character prediction model.
2. The non-destructive egg quality trait prediction method of claim 1, wherein the egg images comprise visible light images, infrared images and hyperspectral images, and the hyperspectral images comprise spectral images of a plurality of wave bands.
3. The method of claim 2, wherein training a pre-defined egg quality trait prediction model based on the image dataset and the genotype dataset comprises:
extracting image features from the image dataset according to a deep convolutional neural network;
coding the DNA sequence in the genotype dataset according to a DNA coding algorithm to obtain a coding result, and extracting DNA sequence characteristics from the coding result according to a deep recurrent neural network;
and fusing the image characteristics and the DNA sequence characteristics to obtain fusion characteristics, and predicting the phenotypic characters of the sample poultry egg from the fusion characteristics according to a full junction layer and a softmax layer.
4. The non-destructive method for predicting the quality and the property of the poultry eggs according to claim 3, wherein the deep convolutional neural network sequentially comprises: convolutional layers, pooling layers, regularization layers, recursive layers, and fully-connected layers.
5. The non-destructive method for predicting the quality traits of poultry eggs according to claim 3, wherein the encoding of the DNA sequences in the genotype dataset according to a DNA encoding algorithm to obtain an encoding result and the extraction of DNA sequence features from the encoding result according to a deep recurrent neural network comprise:
coding the DNA sequence according to Onehot to obtain a first coding result, coding the DNA sequence according to DBE to obtain a second coding result, coding the DNA sequence according to NCP to obtain a third coding result, and coding the DNA sequence according to k _ mer to obtain a fourth coding result;
performing sequence ensemble learning on the first coding result to the fourth coding result to obtain a sequence ensemble learning result;
and inputting the sequence integrated learning result into the deep recurrent neural network to obtain the DNA sequence characteristics extracted by the deep recurrent neural network.
6. The non-destructive poultry egg quality trait prediction method according to claim 5, wherein sequence ensemble learning is performed on the first coding result to the fourth coding result according to XGboost.
7. The method of predicting the quality traits of non-destructive avian eggs according to any one of claims 1 to 6, wherein the phenotypic traits comprise protein content, protein pH, Haught units, lecithin content, egg yolk color, and egg yolk proportion.
8. A non-destructive poultry egg quality trait prediction system is characterized by comprising:
the system comprises a sample acquisition unit, a sample analysis unit and a control unit, wherein the sample acquisition unit is used for acquiring an image data set of a sample egg and a genotype data set of a corresponding laying hen of the sample egg, the image data set comprises a plurality of egg images marked with egg quality phenotypic characters, the genotype data set comprises a DNA sequence of the laying hen, and the DNA sequence is obtained by whole genome DNA re-sequencing or targeted sequencing;
the model training module is used for training a preset egg quality character prediction model according to the image data set and the genotype data set to obtain a trained egg quality character prediction model;
and the model prediction unit is used for predicting the phenotypic character of the target egg from the target image and the target DNA sequence of the target egg according to the trained egg quality character prediction model.
9. An electronic device, characterized in that: comprises at least one control processor and a memory for communicative connection with the at least one control processor; the memory stores instructions executable by the at least one control processor to enable the at least one control processor to perform the non-destructive egg quality trait prediction method of any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: the computer-readable storage medium stores computer-executable instructions for causing a computer to perform the method of predicting a non-destructive avian egg quality trait of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111677548.6A CN114283882B (en) | 2021-12-31 | 2021-12-31 | Non-destructive poultry egg quality character prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111677548.6A CN114283882B (en) | 2021-12-31 | 2021-12-31 | Non-destructive poultry egg quality character prediction method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114283882A CN114283882A (en) | 2022-04-05 |
| CN114283882B true CN114283882B (en) | 2022-08-19 |
Family
ID=80879696
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111677548.6A Active CN114283882B (en) | 2021-12-31 | 2021-12-31 | Non-destructive poultry egg quality character prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114283882B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101930502A (en) * | 2010-09-03 | 2010-12-29 | 深圳华大基因科技有限公司 | Method and system for detection of phenotype genes and analysis of biological information |
| WO2015173435A1 (en) * | 2014-05-16 | 2015-11-19 | Katholieke Universiteit Leuven, KU LEUVEN R&D | Method for predicting a phenotype from a genotype |
| CN105115469A (en) * | 2015-07-29 | 2015-12-02 | 华中农业大学 | Paddy rice spike phenotypic parameter automatic measuring and spike weight predicting method |
| CN105861664A (en) * | 2007-09-26 | 2016-08-17 | 生命科技临床服务实验室公司 | Methods and systems for genomic analysis using ancestral data |
| CN109295052A (en) * | 2018-11-09 | 2019-02-01 | 云南农业大学 | Wooden dipper chicken is without coda gene and detects the method for wooden dipper chicken anury character, primer, kit |
| CN109349222A (en) * | 2018-11-20 | 2019-02-19 | 广西金陵农牧集团有限公司 | A kind of method that molecule auxiliary cultivates short-foot layer of green-shell egg strain |
| CN111291686A (en) * | 2020-02-10 | 2020-06-16 | 中国农业大学 | Method and system for extracting crop root phenotype parameters and judging root phenotype |
| CN111280126A (en) * | 2020-03-30 | 2020-06-16 | 中国农业大学 | Breeding method and application of yellow-feather beard laying hens |
| CN112595675A (en) * | 2020-09-29 | 2021-04-02 | 华中农业大学 | Hyperspectrum-based rice grain quality functional gene analysis method |
| WO2021086595A1 (en) * | 2019-10-31 | 2021-05-06 | Google Llc | Using machine learning-based trait predictions for genetic association discovery |
| CN113228194A (en) * | 2018-10-12 | 2021-08-06 | 人类长寿公司 | Multigroup search engine for comprehensive analysis of cancer genome and clinical data |
| WO2021217138A1 (en) * | 2020-04-24 | 2021-10-28 | TeselaGen Biotechnology Inc. | Method for efficiently optimizing a phenotype with a combination of a generative and a predictive model |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8024128B2 (en) * | 2004-09-07 | 2011-09-20 | Gene Security Network, Inc. | System and method for improving clinical decisions by aggregating, validating and analysing genetic and phenotypic data |
| EP4153782A4 (en) * | 2020-05-22 | 2024-06-26 | Insitro, Inc. | Predicting disease outcomes using machine learned models |
-
2021
- 2021-12-31 CN CN202111677548.6A patent/CN114283882B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105861664A (en) * | 2007-09-26 | 2016-08-17 | 生命科技临床服务实验室公司 | Methods and systems for genomic analysis using ancestral data |
| CN101930502A (en) * | 2010-09-03 | 2010-12-29 | 深圳华大基因科技有限公司 | Method and system for detection of phenotype genes and analysis of biological information |
| WO2015173435A1 (en) * | 2014-05-16 | 2015-11-19 | Katholieke Universiteit Leuven, KU LEUVEN R&D | Method for predicting a phenotype from a genotype |
| CN105115469A (en) * | 2015-07-29 | 2015-12-02 | 华中农业大学 | Paddy rice spike phenotypic parameter automatic measuring and spike weight predicting method |
| CN113228194A (en) * | 2018-10-12 | 2021-08-06 | 人类长寿公司 | Multigroup search engine for comprehensive analysis of cancer genome and clinical data |
| CN109295052A (en) * | 2018-11-09 | 2019-02-01 | 云南农业大学 | Wooden dipper chicken is without coda gene and detects the method for wooden dipper chicken anury character, primer, kit |
| CN109349222A (en) * | 2018-11-20 | 2019-02-19 | 广西金陵农牧集团有限公司 | A kind of method that molecule auxiliary cultivates short-foot layer of green-shell egg strain |
| WO2021086595A1 (en) * | 2019-10-31 | 2021-05-06 | Google Llc | Using machine learning-based trait predictions for genetic association discovery |
| CN111291686A (en) * | 2020-02-10 | 2020-06-16 | 中国农业大学 | Method and system for extracting crop root phenotype parameters and judging root phenotype |
| CN111280126A (en) * | 2020-03-30 | 2020-06-16 | 中国农业大学 | Breeding method and application of yellow-feather beard laying hens |
| WO2021217138A1 (en) * | 2020-04-24 | 2021-10-28 | TeselaGen Biotechnology Inc. | Method for efficiently optimizing a phenotype with a combination of a generative and a predictive model |
| CN112595675A (en) * | 2020-09-29 | 2021-04-02 | 华中农业大学 | Hyperspectrum-based rice grain quality functional gene analysis method |
Non-Patent Citations (6)
| Title |
|---|
| Genetic parameters of egg quality traits on different pedigree layers with special focus on dynamic stiffness;A.E.Blanco 等;《Poultry Science》;20141001;第93卷(第10期);第2457-2463页 * |
| Genome-Wide Association Study and Selection Signatures Detect Genomic Regions Associated with Seed Yield and Oil Quality in Flax;Frank M. You 等;《INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES》;20180831;第19卷(第8期);第1-24页 * |
| Phenotypic Correlation Between Egg Weight and Egg Linear Measurements of the French Broiler Guinea Fowl Raised in the Humid Zone of Nigeria;Dzungwe JT 等;《Current Trends on Biostatistics & 》;20181004;第1卷(第1期);第22-25页 * |
| 基于成像技术的果实表型检测与解析方法研究;李磊;《中国博士学位论文全文数据库 信息科技辑》;20170215(第02期);第I138-168页 * |
| 深度学习在植物表型研究中的应用现状与展望;岑海燕等;《农业工程学报》;20200508(第09期);第1-12页 * |
| 蛋鸡资源群体蛋品质性状的因子分析;曲亮 等;《上海交通大学学报(农业科学版)》;20160630;第34卷(第3期);第1-6页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114283882A (en) | 2022-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Taylor et al. | Guidelines and quantitative standards to improve consistency in cetacean subspecies and species delimitation relying on molecular genetic data | |
| Funk et al. | Harnessing genomics for delineating conservation units | |
| Miao et al. | Automation of leaf counting in maize and sorghum using deep learning | |
| Lu et al. | An automatic splitting method for the adhesive piglets’ gray scale image based on the ellipse shape feature | |
| CN114092378A (en) | Animal health detection method, device, equipment and storage medium | |
| Miele et al. | Deep learning for species identification of modern and fossil rodent molars | |
| CN117274614A (en) | Base recognition method, sequencer and medium based on fluorescence labeling dNTP gene sequencing | |
| CN116580773A (en) | Breeding cross-representation type prediction method and system based on ensemble learning and electronic equipment | |
| Zhang et al. | Comparing deep learning approaches for understanding genotype× phenotype interactions in biomass sorghum | |
| Cheek et al. | Habitat‐linked genetic variation supports microgeographic adaptive divergence in an Island‐endemic bird species | |
| CN116596933B (en) | Base cluster detection method and device, gene sequencer and storage medium | |
| CN114283882B (en) | Non-destructive poultry egg quality character prediction method and system | |
| Barbosa et al. | Wildlife population genomics: Applications and approaches | |
| CN113160886A (en) | Cell type prediction system based on single cell Hi-C data | |
| Mdladla | Landscape genomic approach to investigate genetic adaptation in South African indigenous goat populations. | |
| Benjelloun et al. | Multiple genomic solutions for local adaptation in two closely related species (sheep and goats) facing the same climatic constraints | |
| CN114241409A (en) | Disease early warning system and method based on abnormal excrement and anatomical images of caged chickens | |
| Teixeira et al. | DNA metabarcoding assessment of Neotropical ichthyoplankton communities is marker‐dependent | |
| CN115050437B (en) | Biological big data analysis and disease accurate identification classification prediction system based on algorithm and block chain | |
| Duenk et al. | Predicting the purebred-crossbred genetic correlation from genetic variances within, and covariance between parental lines | |
| Zhang | Deep Learning for Large Scale Plant Phenotyping | |
| CN111754547B (en) | Multi-target automatic identification tracking positioning method for broiler chickens in dim light environment | |
| Bouwman et al. | Classifying true and false reciprocal translocation breakpoints from whole genome sequence data using deep learning | |
| US20230349001A1 (en) | Biomolecular characteristic identification method and system thereof for taiwan tilapia strains | |
| Dwicahyo et al. | Early Detection of Disease in Chicks Using CNN on Bangkok Chicken Health |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |