A reproductive sample information tracking method and system combined with biometric identification

Abstract

The application discloses a reproductive sample information tracking method and system combined with biological feature recognition. The method quantifies the influence of target genes on experimental conditions caused by heredity through an influence coefficient, and to some extent, investigates the correlation between the internal cause of "heredity" and the external cause of "experimental conditions". Although this correlation has no necessary connection with inbreeding degeneration, it is also included in the investigation range considering its negative impact on the experiment. Further, the immune-related gene is one of important factors for ensuring the stability and reliability of experimental data. The application quantifies the influence through an influence degree value, which can effectively quantify the harm and is conducive to realizing the quantization requirement in the reproductive sample information tracking process. It can be seen that the method and system of the application can realize the application of data processing technology for supervision or prediction purposes in the technical field of reproductive sample information tracking.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing for management, supervision or prediction purposes, and in particular to a reproductive sample information tracking method and system combined with biometric identification. BACKGROUND

[0002] The similarity between mouse and human genes is about 90% (about 90% of the 30-40 thousand genes in mice are homologous to humans), especially in key physiological functions such as metabolism and immune system. The gestation period of mice is short (about 20 days), each litter produces 4-12 offspring, and the life cycle is short (2-3 years), so a large number of experimental samples can be obtained quickly. Drug safety testing (such as toxicity and carcinogenicity) relies on mice to obtain core data such as median lethal dose (LD50). More than 90% of preclinical studies of new drugs need to verify the efficacy and side effects through mouse experiments. Mice have become an irreplaceable part of biomedical research due to their high degree of biological similarity to humans, efficient experimental operability, and mature standardization system.

[0003] In scientific research, the stability of the reproductive sample information of mice plays a key role. For example, inbred mice (such as BALB / c and C57BL / 6) are commonly used in research and development, and their genes are highly purified. If the source of reproductive samples is chaotic, it may lead to genetic drift, affecting the consistency of immune response. In a certain adenovirus vaccine experiment, the offspring of the parent mice were not tracked, and the MHC-II class molecule polymorphism difference appeared, resulting in a 30% fluctuation in neutralizing antibody titer. Tracking reproductive sample information can clarify the genetic rules of specific immune characteristics (such as IgG subclass secretion ability) and avoid introducing interference variables due to random mating.

[0004] However, due to the unpredictable inbred degeneration (inbreeding depression) phenomenon, even if the reproductive sample information is tracked, the expression of genetic information is difficult to predict. In addition, the cost of existing reproductive sample information tracking is high, which brings hidden dangers to the realization of reproductive sample information tracking effect.

[0005] Therefore, how to effectively track reproductive sample information has become a problem to be solved.

[0006] The data processing technology suitable for management, supervision or prediction purposes can not only be applied to animal samples such as mice, but also to biological samples such as tomatoes. For example, the publication (announcement) No. CN109840855A, patent title: "a method for predicting whether the tomato yield is reduced in the early reproductive stage" (main classification number: G06Q50 / 02), uses the physiological indicators of tomatoes to establish a model for predicting tomato yield in the early reproductive stage, laying a theoretical foundation for tomato yield reduction warning and high yield optimization conditions. On the one hand, it can be shown that the data processing technology for supervision or prediction purposes has great potential in the field of reproductive sample information tracking related technologies; on the other hand, it can also show that the technical mining in this field has a relatively wide expansion prospect. SUMMARY

[0007] The embodiments of the present application provide a reproductive sample information tracking method and system combined with biological feature recognition to at least partially solve the above technical problems.

[0008] The embodiments of the present application adopt the following technical solutions:

[0009] In a first aspect, the embodiments of the present application provide a reproductive sample information tracking method combined with biological feature recognition, which comprises:

[0010] Data collection is performed on historical risk mice in history to obtain historical data;

[0011] The historical data is used to construct samples and labels corresponding to the samples; the samples represent the biological feature abnormality of the historical risk mice to which they belong, the number of mice contained in the batch to which they belong, the proportion of historical risk mice in the mice of the batch to which they belong, the biological feature abnormality of each mouse contained in the batch to which they belong, the time when the historical risk mice appeared biological feature abnormality, the SNP typing data of the historical risk mice, and the experimental conditions corresponding to the historical risk mice; the labels represent the influence coefficient of the experimental conditions corresponding to the historical risk mice on target genes, and the influence degree of inbreeding degeneration of the historical risk mice on immune-related genes; the influence coefficient is determined based on the performance of the batch to which the historical risk mice belong and its maternal and / or paternal line in the historical experimental conditions; the influence degree is obtained by detecting the genes related to the historical risk mice;

[0012] The samples and the labels are used to train a preset Transformer model to obtain a target model;

[0013] When a suspected risk mouse is found in an experiment, the available data of the suspected risk mouse is input into the target model to obtain the influence coefficient and the influence degree of the suspected risk mouse;

[0014] If the influence coefficient and the influence degree of the suspected risk mouse satisfy a preset first decision condition, it is determined that there is a risk of inbreeding degeneration, and a tracking strategy based on reproductive sample information is executed.

[0015] In an optional embodiment of the present specification, the method further comprises:

[0016] If the influence degree of the suspected risk mouse is greater than a preset degree threshold, it is determined that the first decision condition is satisfied.

[0017] In an optional embodiment of the present specification, the method further comprises:

[0018] The degree threshold in the vaccine experiment scene is less than the degree threshold in other scenes; and / or,

[0019] In the case that the number of mice contained in the batch to which the historical risk mouse belongs is less than a preset number threshold, the judgment for the first decision condition based only on the degree threshold is adopted.

[0020] In an optional embodiment of the present specification, the method further comprises:

[0021] If the influence degree of the suspected risk mouse is not greater than a preset degree threshold, and the influence coefficient is greater than a preset coefficient threshold, it is determined that the first decision condition is satisfied.

[0022] In an optional embodiment of the present specification, the method further comprises:

[0023] The coefficient threshold is positively correlated with the difference between the generation to which the suspected risk mouse belongs and the recommended maximum extension generation, and the degree threshold when the suspected risk mouse satisfies a preset second decision condition is less than the degree threshold when the suspected risk mouse does not satisfy the second decision condition;

[0024] If the mother of the suspected risk mouse, the clone of the mother, the father, or the clone of the father historically has a mutation with respect to the target gene, it is determined that the second decision condition is satisfied.

[0025] In an optional embodiment of the present specification, the method further comprises:

[0026] Reading the RFID chip of the suspected risk mouse to determine the unique identification of the suspected risk mouse;

[0027] Based on the unique identification, data stored in a preset blockchain is searched to obtain SNP typing data of the suspected risk mouse;

[0028] Based on the SNP typing data of the suspected risk mouse, obtain the available data.

[0029] In an alternative embodiment of the present specification, the method further comprises:

[0030] The biological feature abnormality includes at least one of the following: iris texture change, pupil abnormality, eyelid lesion, corneal lesion, sparse hair or alopecia, hair color fading or local whitening, abnormal hair curling, non-traumatic scab or ulcer, decreased skin elasticity, head-body ratio disorder, spinal curvature, walking track deviation from straight line, delayed righting reflex, testis size asymmetry, reduced number of teats, decreased chewing efficiency, smaller fecal particles, and darker urine color.

[0031] In an alternative embodiment of the present specification, the method further comprises:

[0032] The reproductive sample information includes at least one of the following: individual identification, reproductive status, genotype verification, epigenetic marker, SPF level certification, intestinal flora metagenome summary, vertical transmission pathogen screening, operation record, and experimental correlation data.

[0033] In an alternative embodiment of the present specification, the method further comprises:

[0034] After determining the inbreeding degeneration risk, based on the vaginal plug biological feature verification, determine whether the reproductive sample of the SPF level mouse of the suspected risk mouse is contaminated;

[0035] If not, execute the tracking strategy based on the reproductive sample information.

[0036] In a second aspect, the embodiments of the present application also provide a reproductive sample information tracking system combined with biological feature recognition, which comprises:

[0037] A historical data acquisition module configured to collect data for historical risk mice in history to obtain historical data.

[0038] The construction module is configured to: adopt the historical data to construct a sample and a label corresponding to the sample; the sample represents: a biological characteristic abnormality of a historical risk mouse to which the sample belongs, a number of mice contained in a batch to which the sample belongs, a proportion of the historical risk mouse appearing in the mice of the batch, respective biological characteristic abnormalities of the respective mice contained in the batch to which the sample belongs, a time at which the biological characteristic abnormality of the historical risk mouse appears, SNP typing data of the historical risk mouse, and an experimental condition corresponding to the historical risk mouse; and the label represents: an influence coefficient of the experimental condition corresponding to the historical risk mouse on a target gene, and an influence degree of inbreeding degeneration of the historical risk mouse on an immune-related gene; the influence coefficient is determined based on performances of the batch to which the historical risk mouse belongs and its mother line and / or father line in a historical experimental condition; and the influence degree is obtained by performing gene-related detection on the historical risk mouse.

[0039] The training module is configured to: adopt the sample and the label to train a preset Transformer model to obtain a target model.

[0040] The input module is configured to: when a suspected risk mouse is found in an experiment, input available data of the suspected risk mouse into the target model to obtain an influence coefficient and an influence degree of the suspected risk mouse.

[0041] The execution module is configured to: if the influence coefficient and the influence degree of the suspected risk mouse satisfy a preset first decision condition, determine that there is an inbreeding degeneration risk, and execute a tracking strategy based on reproductive sample information.

[0042] In a third aspect, an embodiment of the present application further provides an electronic device, comprising:

[0043] a processor; and

[0044] a memory arranged to store computer executable instructions that, when executed, cause the processor to perform the method steps of the first aspect.

[0045] In a fourth aspect, an embodiment of the present application further provides a computer readable storage medium storing one or more programs, which, when executed by an electronic device comprising a plurality of application programs, cause the electronic device to perform the method steps of the first aspect.

[0046] The above at least one technical solution adopted by the embodiments of the present application can achieve the following beneficial effects:

[0047] The method and system provided in this application first process data, including biometric data, of suspected risk mice based on a target model to obtain an influence coefficient and an influence degree. Then, based on the influence coefficient and influence degree, the presence or absence of inbreeding degeneration is determined, followed by processing of reproductive sample information based on the determination results. This can exclude some sporadic, non-inbreeding degeneration-related suspected risk mice, eliminating the need to track reproductive sample information for all suspected risk mice, effectively saving costs and improving tracking efficiency. Furthermore, the method in this application quantifies the influence of experimental conditions on target genes due to genetics through the influence coefficient, examining to some extent the correlation between the internal factor of "genetics" and the external factor of "experimental conditions." Although this correlation is not necessarily related to inbreeding degeneration, it is included in the scope of consideration considering its negative impact on experiments. Further, immune-related genes are one of the important factors in ensuring the stability and reliability of experimental data. This application quantifies this influence through the influence degree value, effectively quantifying this harm and facilitating the quantitative needs in the process of tracking reproductive sample information. It is evident that the method and system of this application enable the application of data processing technology for monitoring or prediction purposes in the field of reproductive sample information tracking technology. Attached Figure Description

[0048] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0049] Figure 1 A schematic diagram illustrating a method for tracking reproductive sample information that combines biometric identification, provided in an embodiment of this specification.

[0050] Figure 2 This is a schematic diagram of the structure of an electronic device in an embodiment of this specification. Detailed Implementation

[0051] The application will be described in further detail below with specific reference being made to the drawings. Like elements are referred to with like reference numerals throughout the several drawings. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the application. However, those skilled in the art will readily recognize that the various embodiments can be practiced without the specific details given herein. In other instances, well-known methods, procedures and components have not been described in detail so as not to unnecessarily obscure aspects of the application. In addition, the description of the application is not intended to limit the scope of the application, as claimed below, but rather to provide a description of the various embodiments of the application.

[0052] In addition, features, operations or steps described in the specification can be combined in any suitable manner without departing from the scope of the present application. In addition, the various steps or actions in a method described herein can be performed in a different sequence from the one given in any embodiment without departing from the scope of the present application. Thus, the various sequences and steps need not be performed in the order described. All that is required is that the steps be performed in a manner consistent with the description herein. The descriptions and drawings are to be regarded in an illustrative rather than restrictive sense.

[0053] The serial numbers of components in this document, such as "first", "second", etc., are only used to distinguish the described objects, and do not have any sequence or technical meaning. Unless otherwise specified, "connection" and "coupling" in this application include direct and indirect connections (couplings).

[0054] The technical solutions provided by the embodiments of the application will be described in detail below with reference to the drawings.

[0055] As shown in Figure 1 The reproductive sample information tracking method combined with biological feature recognition in the specification includes the following steps:

[0056] S100: Data collection is performed on historical risk mice in history to obtain historical data.

[0057] The mice in the specification are mice that have been or will be used for experiments. The historical mice are mouse individuals that appeared in history. The risk mice are mice that are determined to have inbreeding degeneration.

[0058] The historical data is used for sample construction in the subsequent step. The process of data collection requires data collection for each historical risk mouse. The source of the data mainly has two aspects: one is the RFID chip carried by the mouse; the other is the blockchain. After obtaining the historical data, it can be entered into the designated database for sample construction in the subsequent step. Since there are many experiments based on mice in history, the amount of historical data will not be small enough to construct samples.

[0059] S102: using the historical data, constructing a sample and a label corresponding to the sample.

[0060] The sample and the label in the specification correspond to each other, and one historical risk mouse corresponds to at least one sample. The sample and the label in the specification are used for supervised learning.

[0061] The sample in the specification can be a multi-dimensional vector, which is used to represent: the biological feature abnormality of the historical risk mouse to which it belongs, the number of mice contained in the batch to which it belongs (the same batch of mice may not be used for the same experiment, and even some mice may not participate in the experiment, but as much as possible, they are all counted in to investigate the influence of experimental conditions, and whether the influence on the mouse is realized), the proportion of the historical risk mouse in the mice of the batch to which it belongs (the same batch of mice has high genetic similarity, and in the ideal state, their genes have equal stability. This proportion can reflect the performance of the genes of this batch of mice under a certain experimental condition, which is conducive to reflecting whether the genes of the mouse are suitable for experiments), the biological feature abnormality of each mouse contained in the batch to which it belongs (the difference in biological feature abnormality between different mice in the same batch can reflect the influence of different genes of the mouse under a certain experimental condition. It is also conducive to distinguishing whether the abnormality is caused by inbreeding degeneration or accidental abnormality), the time when the historical risk mouse appears biological feature abnormality (this time can also be represented in a multi-dimensional way, for example, the abnormality occurs at [a day after the mouse is born, b day of participating in the experiment, c day of injecting a certain drug, d day when the abnormality reaches stability and no longer worsens]), SNP typing data of the historical risk mouse (used for genetic influence analysis), and experimental conditions corresponding to the historical risk mouse.

[0062] The biological feature abnormality includes at least one of iris texture change, pupil abnormality (e.g., pupil size asymmetry, pupil light reaction delay, etc.), eyelid lesion (e.g., incomplete eyelid closure, etc.), corneal lesion (e.g., corneal turbidity or vascular hyperplasia, etc.), fur status (e.g., sparse or alopecia, color fading or local whitening, abnormal hair curl, etc.), skin lesion abnormality (e.g., non-traumatic scab or ulcer, skin elasticity decline, etc.), body proportion abnormality (e.g., head-body ratio disorder, spinal curvature, etc.), motion coordination abnormality (e.g., walking trajectory deviates from straight line, righting reflex delay, etc.), external genitalia morphology abnormality (e.g., testis size asymmetry, nipple number reduction, etc.), feeding state abnormality (e.g., reduced chewing efficiency, etc.), excrement abnormality (e.g., fecal particle size becomes smaller, urine color deepens, etc.).

[0063] In the process of inbred line degeneration, the above biological feature abnormalities have the characteristics of being visible to the naked eye and not requiring genetic detection, and can directly reflect the changes in biological features. These features can directly reflect genetic decline and are easy to observe and record. Taking iris texture change as an example, iris texture change can reflect inbred line degeneration. The Cambridge team found that the number of iris crypts in C57BL / 6 mice decreased by 0.3% per generation (significant when F value > 40), and this abnormality can be used for real-time monitoring of genetic drift. In addition, iris texture change can also be used for predictive analysis of immune response. Under certain experimental conditions, biological features are associated with vaccine response, iris vascular density reflects corneal lymphatic vessel hyperplasia, corneal lymphatic vessel hyperplasia reflects lymph node antigen delivery efficiency, and lymph node antigen delivery efficiency may be affected by experimental conditions. Other biological feature abnormalities, such as the number of claw ridge lines, are positively correlated with TLR4 expression and can be used to predict the adjuvant effect of mRNA vaccines.

[0064] In addition to naked eye observation, the technical means for identifying biological feature abnormalities in the related art are applicable to the present specification under the condition that the conditions are allowed.

[0065] Take the iris registration as an example, in an optional embodiment of the present specification, a miniature infrared imaging module (wavelength 850nm LED light source to avoid mouse pupil shrinkage interference, 5 million pixel global shutter CMOS (5μm / pixel resolution), integrated eye axis tracker (compensate for mouse head movement)) is used, and the mouse is allowed to enter a transparent tubular channel (with eyes open in a natural state) with a diameter of 3cm autonomously, and the head is induced to face the imaging window by air flow (pressure <0.1 psi, in line with animal welfare). A light irradiation strategy of visible light first (positioning the eyeball), then near-infrared light (acquiring blood vessel patterns), and then ultraviolet light (enhancing pigment contrast) is used, an improved U-Net algorithm is used to segment the iris region, and a 256-dimensional feature vector (including texture, frequency domain, and topological features) is extracted. The feature vector is bound to the mouse RFID chip number / birth date / parent information.

[0066] However, in some cases, these biometric abnormalities are only occasional phenomena and have nothing to do with inbreeding degeneration. This is what the target model needs to learn in the subsequent steps.

[0067] Among them, the technical means for obtaining SNP typing data in the related art are all applicable to the present specification under the condition of permission. SNP typing data usually contains 50-100 core sites for monitoring genetic drift.

[0068] The experimental conditions are at least one of the following: injection composition and dose, feeding and environmental control (for example, sterile environment: SPF level (specific pathogen free) barrier system, regular disinfection of cages and bedding. Constant temperature and humidity: temperature 22-25℃, humidity 40%-60%, 12 hour light and dark cycle. Standardized feed and water: sterilized feed and filtered drinking water, to avoid environmental factors interfering with experimental results); immunization operation specification (for example, antigen preparation: antigen needs to be emulsified with adjuvant (for example, Freund's complete adjuvant) to ensure stability. Inoculation route and dose: antigen preparation, inoculation route and dose. Immunization program: usually needs multiple immunizations (for example, 0, 14, 28 days), and antibody titer is detected 5-7 days after the last immunization); detection and evaluation index.

[0069] The label in the specification can also be a multi-dimensional vector, used to represent: the influence coefficient of the experimental condition corresponding to the historical risk mouse on the target gene, and the influence degree of inbreeding degeneration of the historical risk mouse on the immune-related gene. In alternative embodiments, the label can be obtained by combining expert experience based on detection data. As a quantitative result, the label can be expressed by a normalized numerical value, whether it is an influence coefficient or an influence degree. As a label, the information it represents should be as true as possible. The value of each dimension of the label can be determined in combination with the corresponding gene detection means. That is, the construction of the label may have a high cost, but it can meet the needs of model training. Through the method in the specification, artificial experience can be quantified to form a computer executable program, which is beneficial to reduce the cost in the long run.

[0070] The influence coefficient in the specification has no unit and is determined based on the performance of the batch to which the historical risk mouse belongs and its maternal and / or paternal line in the historical experimental condition. The influence coefficient is a quantitative determination of the influence of the target gene on the experimental condition due to inheritance, which to some extent investigates the correlation between the internal factor of "inheritance" and the external factor of "experimental condition". The greater the value, the greater the degree of negative influence of the gene carried by the mouse on the experiment. It shows that the mouse carrying this gene is not suitable for the experiment affected by this gene even if it does not occur inbreeding degeneration. It can be understood that: since the experimental condition will cause the influence of the combined action of the target gene and the SNP typing data (one of the reproductive sample information) representing other genes, the mouse carrying the target gene and the SNP typing data representing other genes will be congenitally "degenerated" before the experiment starts, and is not suitable for the experiment (but it cannot be ruled out that the mouse can be used for other experiments unrelated to the target gene).

[0071] For example, in vaccine experiments, "target genes" refer to specific genes that researchers specifically select for modification or study to explore their impact on the immune response induced by vaccines. This concept generally involves two levels: Pathogen-related target genes: These genes encode key antigenic proteins of pathogens (such as the S protein and N protein of the novel coronavirus) for expressing antigens in mice to simulate infection and evaluate the immunogenicity of vaccines. For example, in the development of COVID-19 vaccines, researchers often choose the S protein gene as the target, express the protein through genetic engineering methods, and immunize mice to analyze antibody response titers. Mouse immune-related genes: Through gene editing techniques (such as knockout or overexpression), modify specific immune system genes of mice (such as MHC genes or Toll-like receptor genes) to study the mechanisms of these genes in vaccine responses. For example, knocking out immune checkpoint genes can evaluate their regulatory effects on vaccine protection, or overexpressing specific genes to enhance immune response intensity. Manipulation of these target genes helps optimize vaccine design, including improving antigen expression efficiency or overcoming immune tolerance problems. For example, in the development of genetically engineered antibodies, the selection of target genes directly affects the specificity and affinity of antibodies, providing more precise candidate molecules for vaccines.

[0072] Since the impact coefficient is the comprehensive negative impact of the co-expression of multiple genes, it has certain "innateness" and is more easily perceived. However, this situation is usually rare and is more likely to be discovered in large-scale outbreaks (which can be reflected in the aforementioned "proportion"). More often, it is an accidental event. Mice can carry other genes co-expressed with target genes, which are mostly inherited from their paternal or maternal lines. By analyzing the SNP typing data of their paternal or maternal lines and the experimental conditions, the interaction between this "innateness" and the target gene can be determined to some extent.

[0073] For example, if the paternal or maternal line of a historical risk mouse showed similar "innate degeneration" (e.g., curly hair) in historical experiments (not just for the target gene) and the probability of occurrence was similar, the impact coefficient would decrease. If the batch of historical risk mice showed a significant difference in the probability of "innate degeneration" in historical experiments for the target gene (editing method and target may differ) with changes in experimental conditions, the impact coefficient would decrease. A large number of mice with abnormal biological characteristics appeared in the batch of mice in historical experiments for the target gene, and the impact coefficient would increase.

[0074] The degree of influence in the specification is obtained by performing gene-related detection on the historical risk mice. The greater the value of the degree of influence, the greater the degree of negative influence of inbreeding degeneration on immune-related genes (one of the reproductive sample information). The complete record of mouse reproductive sample information is the core to ensure the traceability of the experiment and the reliability of the data. For example, the genetic expression of immune-related genes can include: parental vaccine response phenotype (e.g., neutralizing antibody titer, Th1 / Th2 bias), histocompatibility antigen (H-2 haplotype) matching record. In addition, the association between the experiment and the immune-related gene is also considered.

[0075] In the related art, the technical means that can be used to implement gene-related detection are all applicable to the specification under the condition of allowing. For example, due to the immune function defect caused by inbreeding degeneration, the T / B cell response of inbred mice is weakened, the amount of antibody production is reduced, and the immunogenicity evaluation in vaccine development is affected. There are corresponding detection means to detect it. For another example, due to the increase of pathogen susceptibility caused by inbreeding degeneration, such as the loss of natural resistance to a specific virus (e.g., LCMV) after inbreeding of SPF level mice, there are corresponding detection means to detect it.

[0076] In the specification, the label is obtained to some extent based on the mouse reproductive sample information. In addition to the foregoing, in an optional embodiment of the specification, the reproductive sample information includes at least one of the following: individual identification, reproductive state, genotype verification, epigenetic marker, SPF level certification, intestinal flora metagenome summary, vertically transmitted pathogen screening, operation record, experimental association data (e.g., vaccine development association data).

[0077] S104: training a preset Transformer model using the sample and the label to obtain a target model.

[0078] The model training in the specification adopts supervised learning. The Transformer model is a deep learning architecture based on self-attention mechanism, which has completely changed the way of processing sequence data. Its core design discards the traditional recurrent neural network, relies on the attention weight matrix to efficiently capture the dependency relationship at any distance, and significantly improves the training efficiency and long sequence modeling capability. The Transformer is composed of an encoder (Encoder) and a decoder (Decoder), and each part is usually stacked with 6-24 layers of structure (the specific number of layers can be adjusted to adapt to the complexity of the task).

[0079] The core mechanism of the Transformer model implicitly reflects several universal principles in information processing:

[0080] 1. Attention mechanism: entropy reduction principle of information filtering. Self-attention aggregates key information by weighting, following the law of information entropy minimization. When the model assigns attention weights to each position in the sequence, it essentially compresses redundant information and focuses on high-value data.

[0081] 2. Parallel architecture: energy-optimal computation. Transformer abandons the serial dependence of RNN, processes sequences in parallel through matrix, and follows the energy minimization path of physical systems: parallel self-attention reduces the computational complexity of long sequences from O(n²) to O(1), reducing energy loss in information transmission. Residual connection maintains the stability of gradient flow, similar to low-impedance path design in circuits.

[0082] 3. Position encoding: embedding of spatiotemporal continuity. Position encoding injects sequence order information, echoing the inseparability of space and time: the position vector generated by the sine function implies periodicity, simulating the wave characteristics in physical systems. Rotational position encoding (RoPE) maintains relative position invariance through complex space rotation, similar to coordinate system transformation in rigid body motion.

[0083] 4. Hierarchical structure: fractal evolution paradigm. The stacked encoder / decoder layers reflect the self-similarity of complex systems: shallow layers capture local syntax (such as word combinations), and deep layers model global semantics (such as logical relationships), forming a fractal feature extraction structure. Feedforward neural networks serve as nonlinear transformers, enabling feature space reconstruction and dimensionality increase.

[0084] 5. Masking mechanism: digitalization of causality. Decoder masking attention strictly follows the law of causality: masking future information ensures that predictions only rely on historical states, simulating the time irreversibility of the physical world. Similar to wave function collapse in quantum measurement, masking converges probability distribution to deterministic output. The model approximates the natural evolution goal of "efficient information processing" through the above mechanisms.

[0085] S106: When a suspected risk mouse is found in the experiment, input the available data of the suspected risk mouse into the target model to obtain the influence coefficient and influence degree of the suspected risk mouse.

[0086] The suspected risk mouse is an abnormal mouse that can be distinguished by biological characteristics, such as the aforementioned biological characteristic abnormalities, which will not be repeated here. Biological characteristics have visible and efficient features. It can be achieved without reading data from the RFID chip. Moreover, the information in the RFID chip is pre-implanted and will not automatically identify and update the biological characteristics of the mouse. That is, the effect of relying solely on the RFID chip is very one-sided. It is difficult for experimenters to rely solely on their experience and naked eye observation to determine whether the abnormality of biological characteristics is inbreeding degeneration or an occasional phenomenon. The target model in the specification is needed to determine it.

[0087] The process of constructing the sample with reference to the historical data has the same or at least partial data dimensions as the available data and the historical data, and the target model can process the available data.

[0088] The generation process of the influence coefficient and the influence degree of the suspected risk mouse is learned by the target model in the training process, and will not be described here.

[0089] In an optional embodiment of the present specification, the process of obtaining available data is: reading the RFID chip of the suspected risk mouse, determining the unique identifier of the suspected risk mouse; based on the unique identifier, searching the data stored in the preset blockchain to obtain the SNP typing data of the suspected risk mouse; and based on the SNP typing data of the suspected risk mouse, obtaining the available data. Since the SNP typing data has a large volume and is difficult to store in the chip, the SNP typing data needs to be stored in another location for reading.

[0090] S108: If the influence coefficient and the influence degree of the suspected risk mouse meet the preset first decision condition, it is determined that there is a risk of inbreeding degeneration, and a tracking strategy based on reproductive sample information is executed.

[0091] The output result of this step is "whether there is a risk of inbreeding degeneration", not "whether there is inbreeding degeneration", and if the result of the judgment is that there is a risk of inbreeding degeneration, the subsequent tracking strategy is executed. If the result of the judgment is that there is no risk of inbreeding degeneration, the subsequent tracking strategy does not need to be executed. This effectively avoids the problem of executing the tracking strategy every time a biological feature is found to be abnormal, and effectively reduces the detection cost.

[0092] In related technologies, the technical means capable of detecting inbreeding degeneration based on reproductive sample information are applicable to the present specification under the condition that the conditions are allowed, and can be used as the tracking strategy in the present specification.

[0093] For example, the tracking strategy can be biochemical marker gene detection, genotyping of biochemical sites on specific chromosomes of inbred animals (such as 14 sites in mice and 11 sites in rats), and determining homozygosity by comparing standard genetic profiles.

[0094] To ensure tracking accuracy, in an optional embodiment of the present specification, after it is determined that there is a risk of inbreeding degeneration, it is determined whether the reproductive sample of the SPF mouse of the suspected risk mouse is contaminated based on the vaginal suppository biological feature verification; if not, the tracking strategy based on the reproductive sample information is executed. If yes, a contamination warning is issued.

[0095] The method and system provided by the present application first process the data of the suspected risk mouse including biological feature data based on a target model to obtain an influence coefficient and an influence degree. Then, the existence of inbreeding degeneration is determined based on the influence coefficient and the influence degree, and then the processing of the reproductive sample information is performed based on the determination result. In this way, some suspected risk mice caused by occasional and non-inbreeding degeneration can be excluded, so that the reproductive sample information of all suspected risk mice does not need to be tracked, the cost can be effectively saved, and the tracking efficiency can be improved. In addition, the method in the present application quantifies the influence of the target gene caused by genetics on the experimental conditions through the influence coefficient, and to some extent, investigates the correlation between the internal factor of "genetics" and the external factor of "experimental conditions". Although this correlation has no necessary connection with inbreeding degeneration, it is also included in the investigation range considering its negative impact on the experiment. Further, the immune-related gene is one of the important factors to ensure the stability and reliability of experimental data, and the present application quantifies this influence through the influence degree, which can effectively quantify this harm and is conducive to realizing the quantization requirement in the reproductive sample information tracking process. It can be seen that the method and system of the present application can realize the application of the data processing technology for supervision or prediction purposes in the technical field related to reproductive sample information tracking.

[0096] The first determination condition will be introduced.

[0097] After obtaining the influence degree and the influence coefficient, in an optional embodiment of the present application, whether the suspected risk mouse has inbreeding degeneration can be determined based on artificial experience. Since the influence degree and the influence coefficient have clear meanings and the causes of the "degeneration" they aim at are also clear, the experimenters can determine whether the "degeneration" is worth considering according to the experimental requirements.

[0098] In another optional embodiment of the present application, only the influence degree is used for determination: if the influence degree of the suspected risk mouse is greater than a preset degree threshold, it is determined that the first determination condition is met. The first determination condition performs well in the case that the number of mice contained in the batch to which the historical risk mouse belongs is less than a preset number threshold (an experience value). Since the number of mice is not large, the number of experimental samples is small, and the abnormality of the biological features exhibited is magnified in the case of small base, resulting in a large error in the determination of "congenital decline".

[0099] The degree threshold in the vaccine experiment scenario is smaller than that in other scenarios, which can improve the sensitivity of detection. For example, when the biological characteristics of a batch of mice are abnormal, the immune characteristics (e.g., MHC haplotype) of the parent mice are traced back through the reproductive sample chain to identify whether the inbreeding degradation is caused. Inbred mice (e.g., BALB / c, C57BL / 6) are often used in vaccine development, and the genes of the inbred mice are highly purified. If the reproductive sample source is confused, genetic drift may occur, which affects the consistency of immune response, and in this case, inbreeding degradation is more alarming.

[0100] In another optional embodiment of the present specification, if the influence degree of the suspected risk mouse is not greater than the preset degree threshold (indicating that although there is a risk, there is no inbreeding degradation), and the influence coefficient is greater than the preset coefficient threshold (indicating that this “degradation phenomenon” is congenital and will also have a negative impact on the experiment), it is determined that the first determination condition is met. The first determination condition in this embodiment is relatively loose, but the factors considered are more comprehensive.

[0101] Further, the coefficient threshold is positively related to the difference between the generation number of the suspected risk mouse and the recommended maximum extension generation number (generally within 10 to 15 generations, and by default, the generation number of the suspected risk mouse is required. The higher the generation number, the greater the risk of genetic instability. The degree threshold when the suspected risk mouse meets the preset second determination condition is less than the degree threshold when the suspected risk mouse does not meet the second determination condition; and if the mother, clone of the mother, father, or clone of the father of the suspected risk mouse has a history of mutation of the target gene, it is determined that the second determination condition is met. The target gene generally refers to a gene that has been compiled, and the previous generation of the suspected risk mouse may not be used for experiments, so there is no target gene. Here, “about the target gene” only refers to the gene that has not been compiled. The mutation of the target gene in the previous generation indicates that the stability of the target gene itself may be a problem, which may be congenital or caused by experimental conditions. In order to ensure the reliability of the experimental results, such mice should be used with caution for certain experiments. The variation of the target gene under certain experimental conditions can also be used to study whether the experimental conditions can be improved.

[0102] Further, the present specification also provides a reproductive sample information tracking system combined with biological characteristic identification, the system comprising:

[0103] The historical data acquisition module is configured to collect data for historical risk mice in history to obtain historical data.

[0104] The construction module is configured to: adopt the historical data to construct a sample and a label corresponding to the sample; the sample represents: biological feature abnormality of a historical risk mouse to which the sample belongs, number of mice contained in a batch to which the sample belongs, proportion of the historical risk mouse appearing in the mice of the batch to which the sample belongs, respective biological feature abnormality of each mouse contained in the batch to which the sample belongs, time when the biological feature abnormality of the historical risk mouse appears, SNP typing data of the historical risk mouse, and experimental condition corresponding to the historical risk mouse; the label represents: influence coefficient of the experimental condition corresponding to the historical risk mouse on a target gene, and influence degree of inbreeding degeneration of the historical risk mouse on an immune-related gene; the influence coefficient is determined based on performance of the batch to which the historical risk mouse belongs and its mother line and / or father line in the historical experimental condition; and the influence degree is obtained by performing gene-related detection on the historical risk mouse.

[0105] The training module is configured to: adopt the sample and the label to train a preset Transformer model to obtain a target model.

[0106] The input module is configured to: when a suspected risk mouse is found in an experiment, input available data of the suspected risk mouse into the target model to obtain the influence coefficient and the influence degree of the suspected risk mouse.

[0107] The execution module is configured to: if the influence coefficient and the influence degree of the suspected risk mouse satisfy a preset first decision condition, determine that there is an inbreeding degeneration risk, and execute a tracking strategy based on reproductive sample information.

[0108] The system can execute the method in any of the preceding embodiments and can obtain the same or similar technical effects, which will not be described here.

[0109] Figure 2 FIG. 1 is a structural schematic diagram of an electronic device according to an embodiment of the present application. Please refer to Figure 2 At the hardware level, the electronic device includes a processor, and optionally further includes an internal bus, a network interface, and a memory. The memory can include a memory such as a random-access memory (RAM), and can also include a non-volatile memory such as at least one disk memory. Of course, the electronic device can also include other hardware required by the business.

[0110] The processor, the network interface and the memory can be connected with each other through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, Figure 2 Only one bidirectional arrow is used to represent the bus, but it does not mean that there is only one bus or only one type of bus.

[0111] The memory is used to store programs. Specifically, the programs can include program codes including computer operation instructions. The memory can include an internal memory and a non-volatile memory, and provide instructions and data for the processor.

[0112] The processor reads the corresponding computer program from the non-volatile memory into the internal memory and then runs, and forms a reproductive sample information tracking system combined with biological feature recognition at a logical level. The processor executes the programs stored in the memory, and is specifically used for executing any one of the above reproductive sample information tracking methods combined with biological feature recognition.

[0113] The above as described in the present application Figure 1The reproductive sample information tracking method with biometric recognition disclosed by the embodiment can be applied to a processor or implemented by the processor. The processor can be an integrated circuit chip with signal processing capability. In the implementation process, each step of the method can be completed by integrated logic circuits of hardware in the processor or instructions in the form of software. The processor can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; or a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. Each method, step and logic block disclosed in the embodiment of the present application can be implemented or executed. The general processor can be a microprocessor or any conventional processor. The steps of the method disclosed in combination with the embodiment of the present application can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register or other mature storage medium in the art. The storage medium is located in the memory, and the processor reads the information in the memory and combines the hardware to complete the steps of the method.

[0114] The electronic device can further execute Figure 1 a reproductive sample information tracking method with biometric recognition, and implement Figure 1 the functions of the embodiment, which will not be described here in detail.

[0115] The embodiment of the present application further proposes a computer readable storage medium storing one or more programs, the one or more programs including instructions for executing any of the aforementioned reproductive sample information tracking methods with biometric recognition when executed by an electronic device including a plurality of applications.

[0116] Those skilled in the art will appreciate that embodiments of the present application can be readily used as software, hardware, or a combination of software and hardware. In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0117] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 a system to perform the function specified in the flowchart block or blocks.

[0118] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 a system to perform the function specified in the flowchart block or blocks.

[0119] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 a system to perform the function specified in the flowchart block or blocks.

[0120] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0121] The memory can include non-persistent memory and / or persistent memory, such as flash memory, or a readonly memory (ROM). The memory is an example of computer readable media.

[0122] Computer-readable media includes permanent and non-permanent, movable and non-movable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer-readable media does not include transitory media such as modulated data signals and carriers.

[0123] It should also be noted that the terms "comprising", "containing", or any other variant thereof are intended to cover non-exclusive inclusions, such that a process, method, article or apparatus that comprises a list of elements does not only include those elements, but also other elements not explicitly listed, or other elements inherent to such a process, method, article or apparatus. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article or apparatus that includes the element.

[0124] Those skilled in the art will appreciate that embodiments of the present application can be provided as methods, systems or computer program products. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) containing computer usable program code.

[0125] The above only describes the embodiments of the present application and is not intended to limit the present application. Those skilled in the art can make various changes and modifications to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the scope of the claims of the present application.

Claims

1. A reproductive sample information tracking method incorporating biometric identification, characterized by, The method comprises: Data collection is performed on historical risk mice in history to obtain historical data; Using the historical data, samples and labels corresponding to the samples are constructed; the samples represent: biological feature abnormality of the historical risk mice to which the samples belong, number of mice contained in the batch to which the samples belong, proportion of historical risk mice in the mice of the batch to which the samples belong, biological feature abnormality of each mouse contained in the batch to which the samples belong, time when the historical risk mice appeared biological feature abnormality, SNP typing data of the historical risk mice, and experimental conditions corresponding to the historical risk mice; the labels represent: influence coefficient of the experimental conditions corresponding to the historical risk mice on target genes, and influence degree of inbreeding degeneration of the historical risk mice on immune-related genes; the influence coefficient is determined based on performance of the batch to which the historical risk mice belong and its maternal and / or paternal line in historical experimental conditions; the influence degree is obtained by performing gene-related detection on the historical risk mice; The preset Transformer model is trained using the samples and the labels to obtain a target model; When a suspected risk mouse is found in an experiment, available data of the suspected risk mouse is input into the target model to obtain the influence coefficient and the influence degree of the suspected risk mouse, wherein the available data of the suspected risk mouse is biological feature abnormality of the historical risk mouse to which the suspected risk mouse belongs, number of mice contained in the batch to which the suspected risk mouse belongs, proportion of historical risk mice in the mice of the batch to which the suspected risk mouse belongs, biological feature abnormality of each mouse contained in the batch to which the suspected risk mouse belongs, time when the historical risk mouse appeared biological feature abnormality, SNP typing data of the historical risk mouse, and experimental conditions corresponding to the historical risk mouse; If the influence coefficient and the influence degree of the suspected risk mouse satisfy a preset first decision condition, it is determined that there is an inbreeding degeneration risk, and a tracking strategy based on reproductive sample information is executed.

2. The method of claim 1, wherein, The method further comprises: If the influence degree of the suspected risk mouse is greater than a preset degree threshold, it is determined that the first decision condition is satisfied.

3. The method of claim 2, wherein, The method further comprises: In a vaccine experiment scenario, the degree threshold is less than the degree threshold in other scenarios; and / or, In a case where the number of mice contained in the batch to which the suspected risk mouse belongs is less than a preset number threshold, a judgment on the first decision condition based only on the degree threshold is adopted.

4. The method of claim 1, wherein, The method further comprises: If the influence degree of the suspected risk mouse is not greater than a preset degree threshold, and the influence coefficient is greater than a preset coefficient threshold, it is determined that the first decision condition is satisfied.

5. The method of claim 4, wherein, The method further comprises: The coefficient threshold is positively correlated with a difference between the generation number of the suspected risk mouse and a recommended maximum continuous generation number, and the degree threshold when the suspected risk mouse satisfies a preset second decision condition is less than the degree threshold when the suspected risk mouse does not satisfy the second decision condition. In a case where the mother of the suspected risk mouse, a clone of the mother, the father, or a clone of the father has a mutation with respect to the target gene in the past, it is determined that the second determination condition is satisfied.

6. The method of claim 1, wherein, The method further includes: reading an RFID chip of the suspected risk mouse to determine a unique identifier of the suspected risk mouse; based on the unique identifier, searching data stored in a preset blockchain to obtain SNP typing data of the suspected risk mouse; based on the SNP typing data of the suspected risk mouse, obtaining the available data.

7. The method of claim 1, wherein, The method further includes: The biological feature abnormality includes at least one of the following: iris texture change, pupil abnormality, eyelid lesion, corneal lesion, hair thinning or alopecia, hair color fading or local whitening, abnormal hair curling, non-traumatic scab or ulcer, decreased skin elasticity, head-body ratio disorder, spinal curvature, walking track deviation from a straight line, delayed righting reflex, testis size asymmetry, reduced number of teats, decreased chewing efficiency, smaller fecal particles, and darker urine color.

8. The method of claim 1, wherein, The method further includes: The reproductive sample information includes at least one of the following: individual identification, reproductive state, genotype verification, epigenetic marker, SPF level certification, intestinal flora metagenome summary, vertical transmission pathogen screening, operation record, and experimental correlation data.

9. The method of claim 1, wherein, The method further includes: After determining that there is a risk of inbreeding degeneration, based on the vaginal plug biological feature verification, determining whether the reproductive sample of the suspected risk mouse is contaminated; if not, performing a tracking strategy based on the reproductive sample information.

10. A reproductive sample information tracking system incorporating biometric identification, characterized by, The system includes: a historical data acquisition module configured to collect data of historical risk mice in the past to obtain historical data; a construction module configured to construct samples and labels corresponding to the samples using the historical data; the samples represent biological feature abnormalities of the historical risk mice to which the samples belong, the number of mice included in the batches to which the samples belong, the proportion of the historical risk mice among the mice in the batches to which the samples belong, the respective biological feature abnormalities of each mouse included in the batches to which the samples belong, the time at which the biological feature abnormalities of the historical risk mice occurred, SNP typing data of the historical risk mice, and experimental conditions corresponding to the historical risk mice; the labels represent influence coefficients of the experimental conditions corresponding to the historical risk mice on target genes and influence degrees of inbreeding degeneration of the historical risk mice on immune-related genes; the influence coefficients are determined based on the performance of the batches to which the historical risk mice belong and their maternal and / or paternal lines in historical experimental conditions; the influence degrees are obtained by performing gene-related detection on the historical risk mice; a training module configured to train a preset Transformer model using the samples and the labels to obtain a target model; and a determination module configured to determine, based on the target model, whether the second determination condition is satisfied. The input module is configured to: when a suspected risk mouse is found in an experiment, input available data of the suspected risk mouse into the target model to obtain an influence coefficient and an influence degree of the suspected risk mouse, wherein the available data of the suspected risk mouse comprises biological feature abnormality of a historical risk mouse to which the suspected risk mouse belongs, a number of mice contained in a batch to which the suspected risk mouse belongs, a proportion of the historical risk mouse in the mice of the batch, respective biological feature abnormalities of the mice contained in the batch, a time when the historical risk mouse appears the biological feature abnormality, SNP typing data of the historical risk mouse, and an experimental condition corresponding to the historical risk mouse; The execution module is configured to: if the influence coefficient and the influence degree of the suspected risk mouse satisfy a preset first decision condition, determine that there is a risk of inbreeding degeneration, and execute a tracking strategy based on reproductive sample information.

Images