Gene reproducibility calculation device, gene reproducibility calculation method, and computer program

Abstract

According to embodiments of the present invention, the device and method for calculating gene reproducibility can be provided, the device and method predicting drug reactivity related to a gene in an organoid by using the degree of similarity between an expression pattern of the gene in the organoid and an expression pattern of the gene in a tissue cell. In addition, according to embodiments of the present disclosure, the device and method for calculating gene reproducibility can be provided, the device and method determining an organoid suitable for drug development related to a gene by using an expression pattern of the gene in the organoid.

Description

Genetic reproducibility calculation device, genetic reproducibility calculation method and computer program

[0001] The present disclosure relates to a genetic reproducibility calculation device, a genetic reproducibility calculation method, and a computer program, and more specifically, is characterized in that the genetic reproducibility level of an organoid is calculated using a correlation coefficient value between gene expression levels in tissue cells and organoids.

[0002] Current translational medicine research is widely utilizing organoids to mimic human disease states and test drug responses. This approach plays a particularly important role in cancer research and drug development. Because organoids are composed of cells derived directly from actual human tissue, drug testing results in organoids can replicate drug responses in actual patients.

[0003] If organoids can replicate the complex gene expression patterns and drug response mechanisms of actual tissues, drug responses can be recreated using organoids. Conventional methods have been able to measure reproducibility in organoids by directly comparing RNA sequencing data extracted from tissues with RNA sequencing data extracted from organoids.

[0004] However, this method only measures the overall similarity of gene expression patterns between organoids and tissues, and does not include detailed measurements of how well each gene reproduces specific expression patterns in the tissue. Furthermore, this method cannot statistically assess the reproducibility of specific gene expression patterns in organoids. Furthermore, this method fails to account for the issue that some genes do not play important roles in organoids. In other words, as described above, conventional methods only measure the similarity between genomic data in organoids, but cannot measure reproducibility.

[0005] Therefore, there is an emerging need to more accurately assess how well the expression patterns of each gene within the tissue are reproduced in organoids and how significant this is.

[0006] To address the aforementioned issues, the degree of similarity between the expression patterns of genes in organoids and those in tissue cells can be used to quantitatively measure drug responsiveness associated with genes in organoids.

[0007] Gene expression patterns in organoids can be used to determine which organoids are suitable for gene-related drug development.

[0008] The similarity of gene expression patterns in organoids compared to tissues can be statistically evaluated.

[0009] According to embodiments of the present disclosure, a method comprises: a step in which a gene reproducibility calculation device receives information about a first gene; a step in which the gene reproducibility calculation device searches for first tissue cells in which a first gene is expressed and first organoids corresponding to the first tissue cells, sets a combination pair including one of the first tissue cells and one of the first organoids, and calculates first correlation coefficients between expression levels of the first gene for each combination pair using a first method; a step in which the gene reproducibility calculation device calculates, using a second method different from the first method, second correlation coefficients between expression levels of the first gene for each combination pair using a second method; a step in which the gene reproducibility calculation device sets random combination pairs for the first tissue cells and the first organoids, calculates first random correlation coefficients between expression levels of the first gene for the random combination pairs using the first method, and calculates a probability value of the first correlation coefficient in a distribution of the first random correlation coefficients; The above-mentioned gene reproducibility calculation device may include a step of calculating second random correlation coefficients between the expression levels of the first gene for the random combination pairs using a second method, and calculating a probability value of the second correlation coefficient in a distribution of the second random correlation coefficients; and a step of calculating the reproducibility level of the first gene based on the probability value of the first correlation coefficient or the probability value of the second correlation coefficient.

[0010] The correlation coefficient calculated by the first method above may be a Pearson correlation coefficient.

[0011] The correlation coefficient calculated by the second method may be a Spearman correlation coefficient.

[0012] According to the present embodiment, the genetic reproducibility calculation device may further include a step of generating data on the reproducibility level as output data and transmitting the data to another device.

[0013] The step of calculating the reproduction level of the first gene may determine the presence or absence of the reproduction level of the first gene based on whether the probability value of the first correlation coefficient is greater than or equal to a first threshold value and whether the probability value of the second correlation coefficient is greater than or equal to a second threshold value.

[0014] The above first or second threshold value may be a value output from a learning device that learns using machine learning or deep learning.

[0015] The above learning device may be capable of learning by inputting genetic information, correlation coefficients calculated by comparing gene expression levels, and probability values ​​of correlation coefficients.

[0016] A genetic reproducibility calculation device according to embodiments of the present disclosure comprises: a data input unit that receives information about a first gene, searches for first tissue cells in which the first gene is expressed, and first organoids corresponding to the first tissue cells, and sets a combination pair including one of the first tissue cells and one of the first organoids; a first correlation coefficient calculation unit that calculates first correlation coefficients between expression levels of the first gene for each combination pair using a first method; A second method different from the first method, comprising a second correlation coefficient calculation unit for calculating second correlation coefficients between expression levels of the first gene for each combination pair using the second method; wherein the first correlation coefficient calculation unit sets random combination pairs for the first tissue cells and the first organoids, calculates first random correlation coefficients between expression levels of the first gene for the random combination pairs using the first method, and calculates a probability value of the first correlation coefficient in a distribution of the first random correlation coefficients; and the second correlation coefficient calculation unit calculates second random correlation coefficients between expression levels of the first gene for the random combination pairs using the second method, and calculates a probability value of the second correlation coefficient in a distribution of the second random correlation coefficients; and may further include a reproduction level calculation unit for calculating a reproduction level of the first gene based on the probability value of the first correlation coefficient or the probability value of the second correlation coefficient.

[0017] A computer program according to an embodiment of the present invention can be stored in a medium to execute any one of the methods according to an embodiment of the present invention using a computer.

[0018] In addition, other methods for implementing the present invention, other systems, and computer-readable recording media recording a computer program for executing the above methods are further provided.

[0019] Other aspects, features and advantages other than those described above will become apparent from the following drawings, claims and detailed description of the invention.

[0020] According to embodiments of the present invention, a device and method for calculating gene reproducibility can be provided that can quantitatively measure drug responsiveness related to a gene in an organoid by utilizing the degree of similarity between the expression pattern of a gene in an organoid and the expression pattern of a gene in a tissue cell.

[0021] A genetic reproducibility calculation device and method can be provided that can determine an organoid suitable for drug development related to a gene by using the expression pattern of a gene in the organoid.

[0022] A device and method for calculating genetic reproducibility can be provided that can statistically evaluate the similarity of the gene expression pattern of an organoid compared to an organization.

[0023] FIG. 1 is a block diagram of a genetic reproducibility calculation device according to embodiments of the present disclosure.

[0024] FIG. 2 is a drawing of a model included in a reproducibility calculation device according to embodiments of the present disclosure.

[0025] FIG. 3 is a drawing of a learning device (200) that calculates a threshold value for a reproduction level in a genetic reproducibility calculation device (100).

[0026] Figure 4 is a table of the first and second correlation coefficients for each gene derived from the patient's lung cancer tissue and organoids derived from the lung cancer tissue.

[0027] Figure 5 is a diagram showing the coordinates of combination pairs between lung cancer tissues and organoids.

[0028] Figure 6 is an exemplary distribution diagram illustrating first random correlation coefficients of random pairs between lung cancer tissue of a patient and organoids derived from lung cancer tissue.

[0029] Figure 7 is an exemplary distribution diagram illustrating second random correlation coefficients of random pairs between lung cancer tissue of a patient and organoids derived from lung cancer tissue.

[0030] Figure 8 is a flowchart of a method for calculating genetic reproducibility according to embodiments of the present disclosure.

[0031] The configuration and operation of the present invention will be described in detail with reference to embodiments of the present invention illustrated in the attached drawings below.

[0032] The present invention is capable of various modifications and embodiments. Specific embodiments are illustrated in the drawings and described in detail in the detailed description. The effects and features of the present invention, as well as the methods for achieving them, will become clearer with reference to the embodiments described in detail below, along with the drawings. However, the present invention is not limited to the embodiments disclosed below and can be implemented in various forms.

[0033] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings. When describing with reference to the drawings, identical or corresponding components are given the same drawing reference numerals, and redundant descriptions thereof will be omitted.

[0034] Hereinafter, the term “upper” or “upper” may include not only things that are directly above in contact, but also things that are above in a non-contact manner.

[0035] In the following examples, the terms first, second, etc. are not used in a limiting sense, but are used for the purpose of distinguishing one component from another.

[0036] In the examples below, singular expressions include plural expressions unless the context clearly indicates otherwise.

[0037] In the following examples, terms such as “include” or “have” mean that a feature or component described in the specification is present, and do not preclude the possibility that one or more other features or components may be added.

[0038] For convenience of explanation, the sizes of components in the drawings may be exaggerated or reduced. For example, the sizes and thicknesses of each component shown in the drawings are arbitrarily indicated for convenience of explanation, and thus the present invention is not necessarily limited to what is shown.

[0039] Additionally, terms such as “…part”, “…area”, etc. described in this specification may mean a unit that processes at least one function or operation.

[0040] Here, organoids refer to three-dimensional cell aggregates formed through self-renewal and self-organization from adult stem cells, embryonic stem cells, or induced pluripotent stem cells. Organoids can contain specific cells of model organs by re-aggregating and recombining cells using three-dimensional culture techniques.

[0041] Here, tissue cells refer to cells present within an organ in biology. In this specification, tissue cells are collected from a subject (human, animal, plant, etc.) and may include cancer cells associated with disease.

[0042] Here, the correlation coefficient refers to the relationship between two variables. If the correlation coefficient is a positive (+) value, it means a positive correlation, if the correlation coefficient is 0, it means no correlation, and if the correlation coefficient is a negative (-) value, it means a negative correlation. In this specification, if the correlation coefficient is a predetermined threshold value, for example, 0.9 or higher, it can be determined that there is similarity between organoids and tissue cells.

[0043] FIG. 1 is a block diagram of a genetic reproducibility calculation device according to embodiments of the present disclosure.

[0044] The gene reproducibility calculation device (100) can calculate the reproducibility level for the expression pattern of a gene for an organoid. The gene reproducibility calculation device (100) can calculate the degree to which the expression pattern of a first gene is reproducible in an organoid, in response to input data of a first gene. The gene reproducibility calculation device (100) can calculate the reproducibility level by comparing the expression pattern of the first gene in the organoid with the expression pattern of the first gene in tissue cells. The reproducibility level refers to a value quantified according to a predetermined standard. The standard applied to the reproducibility level can be changed.

[0045] The expression pattern of the first gene in the organoid can be determined by comparing the expression-related data of the first gene in tissue cells with the expression-related data of the first gene in the organoid.

[0046] More specifically, the genetic reproducibility calculation device (100) can calculate a correlation coefficient between the expression level of a gene in a tissue cell and the expression level of a gene in an organoid, and calculate a reproduction level in the organoid based on the correlation coefficient. When the correlation coefficient is closer to 1 or higher than a predetermined maximum threshold value, the reproduction level of the organoid can be determined to be high. When the correlation coefficient is closer to -1 or lower than a predetermined minimum threshold value, the reproduction information of the organoid can be determined to be low. The maximum threshold value or the minimum threshold value can be changed.

[0047] The genetic reproducibility calculation device (100) searches for tissue cells and organoids in relation to the input first gene, and can utilize the distribution of correlation coefficients calculated between the tissue cells and organoids. The distribution of correlation coefficients (see FIG. 5) refers to a graph that distributes correlation coefficients of multiple combinations between tissue cells and organoids according to the magnitude of the values.

[0048] In an optional embodiment, the genetic reproducibility calculation device (100) can calculate the reproducibility level of organoids with respect to all genes of tissue cells to be reproduced. The genetic reproducibility calculation device (100) can receive information about tissue cells and calculate the reproducibility level of organoids with respect to one or more genes contained in the tissue cells.

[0049] The genetic reproducibility calculation device (100) calculates a first correlation coefficient between a first organoid and a first tissue cell for which a reproducibility level is desired, and statistically calculates the reproducibility level in the first organoid based on the position value of the first correlation coefficient in the distribution of correlation coefficients.

[0050] The genetic reproducibility calculation device (100) can calculate a plurality of correlation coefficients between the first organoid and the first tissue cells using various correlation coefficient calculation methods. In addition, the genetic reproducibility calculation device (100) can generate a distribution of correlation coefficients using various correlation coefficient calculation methods. For example, the genetic reproducibility calculation device (100) can calculate the position values ​​of the correlation coefficients calculated using the first method and the position values ​​of the correlation coefficients calculated using the second method, respectively.

[0051] In another embodiment, the correlation coefficient may be calculated by taking into account the order of expression of genes other than the expression amount of genes in organoids and tissue cells.

[0052] The genetic reproducibility calculation device (100) can generate matching information (information on combination pairs) that randomly connects combinations of multiple tissue cells and multiple organoids to generate a distribution map of correlation coefficients, and can calculate a correlation coefficient of gene expression levels between combinations of tissue cells and organoids according to each of the matching information. The genetic reproducibility calculation device (100) can generate a distribution map of correlation coefficients, which is a distribution map of correlation coefficients calculated from randomly connected matching information, as shown in FIG. 5. Here, the distribution map of correlation coefficients may be a null distribution.

[0053] The genetic reproducibility calculation device (100) can calculate the position value of the first correlation coefficient between a tissue and an organoid from a distribution of correlation coefficients using one or more calculation methods. For example, a method such as Pearson or Spearman can be used.

[0054] Through this, the gene reproducibility calculation device (100) can predict drug responsiveness related to a gene in an organoid by utilizing the degree of similarity between the expression pattern of a gene in an organoid and the expression pattern of a gene in a tissue cell. The gene reproducibility calculation device (100) can determine an organoid suitable for drug development related to a gene by utilizing the expression pattern of a gene in an organoid.

[0055] The genetic reproducibility calculation device (100) may include a data input unit (110), a first correlation coefficient calculation unit (120), a second correlation coefficient calculation unit (130), and a reproducibility level calculation unit (140).

[0056] The data input unit (110) can receive information about a first gene. The data input unit (110) can receive information about tissue cells and extract information about any one of the first genes included in the tissue cells.

[0057] The data input unit (110) can search for tissue cells (hereinafter, a set of tissues) in which the first gene is expressed and organoids (hereinafter, a set of organoids) corresponding to the tissue cells.

[0058] The data input unit (110) can generate combination pairs (hereinafter, a set of combination pairs) between one or more tissue cells included in a set of tissues and one or more organoids included in a set of organoids. The set of tissues may be one tissue, and multiple organoids derived from one tissue may be included in the set of organoids, but is not limited thereto. One combination pair included in the set of combination pairs is referred to as a first combination pair. The first combination pair may include a first tissue cell and a first organoid that correspond to each other. The first tissue cell may be included in the set of tissues, and the first organoid may be included in the set of organoids.

[0059] The first correlation coefficient calculation unit (120) can calculate the first correlation coefficient based on the relationship between the gene expression levels of the first tissue cells included in the first combination pair and the gene expression levels of the first organoid. The first correlation coefficient calculation unit (120) can display the gene expression levels in the tissue cells of each combination pair and the gene expression levels in the organoid as two-dimensional coordinates (see FIG. 5). The first correlation coefficient can be calculated based on the relationship between the coordinates corresponding to the combination pair. For example, the relationship y=ax can be derived between the coordinates and the value a can be derived as the first correlation coefficient. As a result of deriving the relationship between the coordinates of FIG. 5, the value a is calculated as 0.847738.

[0060] The first correlation coefficient calculation unit (120) can create combination pairs by configuring a set of tissue cells and a set of organoids for each gene, and calculate the relationship between the combination pairs as a first correlation coefficient. As illustrated in Fig. 4, the first correlation coefficients can be calculated for each gene (gene ID) as 0.848559, 0.847738, 0.821631, 0.800514, 0.777551, etc.

[0061] Here, gene expression levels can be measured based on RNA sequencing data for organoids or tissue cells. Furthermore, the first correlation coefficient can be viewed as a numerical indicator of how similar the gene expression patterns are between tissues and organoids.

[0062] The first correlation coefficient can be calculated by the mathematical formula below.

[0063]

[0064] Here, X may be the expression level of the first gene of the tissue cell, and Y may be the expression level of the first gene of the organoid.

[0065] The first correlation coefficient calculation unit (120) can generate n random combination pairs by randomly matching one or more tissue cells included in the set of tissue cells and one or more organoids included in the set of organoids in a one-to-one relationship. Through the above process, the first correlation coefficient calculation unit (120) can generate first random combination pairs, second random combination pairs, third random combination pairs, … n-th random combination pairs. The first to n-th random combination pairs can be generated without overlapping each other. n is a natural number, and can refer to the number of ways to create random combination pairs without overlapping between the set of tissue cells and the set of organoids.

[0066] The first correlation coefficient calculation unit (120) can calculate correlation coefficients between the gene expression levels of tissue cells of the first random combination pairs and the gene expression levels of the organoids. In this way, the correlation coefficient calculation unit (120) can calculate correlation coefficients based on the relationship between the gene expression levels of tissue cells of the second to n-th random combination pairs and the gene expression levels of the organoids. The first correlation coefficient calculation unit (120) can set the correlation coefficients calculated by the first correlation coefficient method for the first to n-th random combination pairs as first random correlation coefficients. One correlation coefficient is calculated for the first random combination pair, and n correlation coefficients can be calculated for the first to n-th random combination pairs. The first correlation coefficient calculation unit (120) can generate a first correlation coefficient distribution by arranging the first random correlation coefficients in order of size. The first correlation coefficient distribution can be generated by the 3-sigma rule. The first correlation coefficient calculation unit (120) can generate a probability value of the first correlation coefficient in the first correlation coefficient distribution. More specifically, the probability value of the first correlation coefficient may be a statistical significance value (P) of the first correlation coefficient. For example, if the first correlation coefficient of the organoid calculated for the first gene is 0.28, the position of the 0.28 value corresponds to the mean + standard deviation (-0.00008258 + 0.20836321) ~ mean + 2X standard deviation (-0.00008258 + 2X0.20836321), and the probability value of the second correlation coefficient is calculated accordingly.

[0067] The second correlation coefficient calculation unit (130) can calculate the second correlation coefficient based on the relationship between the gene expression levels of the first tissue cells included in the first combination pair and the gene expression levels of the first organoid. The second correlation coefficient may be calculated using a different correlation coefficient method than the first correlation coefficient. For example, the first correlation coefficient may be a Pearson correlation coefficient, and the second correlation coefficient may be a Spearman correlation coefficient. In addition, the second correlation coefficient may be viewed as an indicator that numerically represents how similar the gene expression patterns are between the tissue and the organoid.

[0068] The second correlation coefficient calculation unit (130) can display the gene expression levels in the tissue cells of each combination pair and the gene expression levels in the organoid as two-dimensional coordinates. The first correlation coefficient can be calculated based on the relationship between the coordinates corresponding to the combination pair. For example, the relationship y=bx can be derived between the coordinates, and the b value can be derived as the second correlation coefficient. As a result of deriving the relationship between the coordinates of FIG. 5 using the second method, the b value can be calculated as 0.752174.

[0069] The second correlation coefficient calculation unit (130) can create combination pairs by configuring a set of tissue cells and a set of organoids for each gene, and calculate the relationship between the combination pairs as a second correlation coefficient. As illustrated in Fig. 4, the second correlation coefficients can be calculated for each gene (gene ID) as 0.786957, 0.752174, 0.512174, 0.658261, 0.769565, etc.

[0070] The second correlation coefficient calculation unit (130) can generate n random combination pairs by randomly matching one or more tissue cells included in the set of tissue cells and one or more organoids included in the set of organoids in a one-to-one relationship. Through the above process, the second correlation coefficient calculation unit (130) can generate first random combination pairs, second random combination pairs, third random combination pairs, … n-th random combination pairs. The first to n-th random combination pairs can be generated without overlapping each other.

[0071] The second correlation coefficient calculation unit (130) can calculate one correlation coefficient based on the relationship between the gene expression levels of the tissue cells of the first random combination pairs and the gene expression levels of the organoid. In this way, the second correlation coefficient calculation unit (130) can calculate correlation coefficients between the gene expression levels of the tissue cells of the second to n-th random combination pairs and the gene expression levels of the organoid. The second correlation coefficient calculation unit (130) can set n correlation coefficients calculated by the second correlation coefficient method for the first to n-th random combination pairs as second random correlation coefficients. The second correlation coefficient calculation unit (130) can generate a second correlation coefficient distribution considering the second random correlation coefficients. The second correlation coefficient distribution can be generated by the 3-sigma rule. The second correlation coefficient calculation unit (130) can generate a probability value of the second correlation coefficient in the second correlation coefficient distribution. More specifically, the probability value of the second correlation coefficient may be the statistical significance value (P) of the second correlation coefficient. For example, if the second correlation coefficient of the organoid calculated for the first gene is 0.25, the position value of the 0.25 value corresponds to the mean + standard deviation (-0.00081206 + 0.20836851) ~ mean + 2X standard deviation (-0.00081206 + 2X0.20836851), and the probability value of the second correlation coefficient is calculated accordingly.

[0072] The reproduction level calculation unit (140) can calculate the reproduction level of the first gene in the organoid by considering the first probability value corresponding to the position value of the first correlation coefficient and the second probability value corresponding to the position value of the second correlation coefficient. At this time, the reproduction level calculation unit (140) can calculate the statistical significance value for the first probability value and the second probability value using the Stouffer method. The reproduction level of the first gene in the organoid can be calculated based on the statistical significance value calculated in this way. The reproduction level can be determined to be higher as the first and / or second probability values ​​are larger.

[0073] The reproducibility level can be calculated according to a predetermined rule. For example, it can be determined using a table or relational expression that defines the reproducibility level in response to the first and / or second probability values.

[0074] The output data generation unit (150) can generate output data regarding the reproduction level of the first gene in the organoid. The output data can include at least one of a distribution of the first and / or second correlation coefficients, a first and / or second probability value, and a first and / or second correlation coefficient. The output data can include data visually representing the reproduction level of the first gene in the organoid. The output data can include the presence or absence of reproduction (Yes / no) or whether the reproduction level is high or low in comparison with a predetermined threshold. The output data generation unit (150) can generate information regarding an organoid having a reproduction level higher than a predetermined threshold as output data.

[0075] For example, organoids having first and / or second probability values ​​greater than or equal to a threshold of 0.75 may be output as having a high level of reproducibility. Information about organoids having first and / or second probability values ​​greater than or equal to a threshold of, for example, 0.8, input by the user may be output.

[0076] The genetic reproducibility calculation device (100) can be configured to charge a fee corresponding to the provided output data. The fee may be determined in proportion to the number of organoids included in the output data. The organoids included in the output data may have a reproducibility level greater than or equal to a predetermined threshold.

[0077] FIG. 2 is a drawing of a model included in a reproducibility calculation device according to embodiments of the present disclosure.

[0078] M1 of the genetic reproducibility calculation device (100) can search for a set of tissues (TS1) in which a first gene is expressed and a set of organoids (OS1) in which the first gene is expressed. M1 can generate a first combination pair (CP1) including tissue cells of TS1 and organoids of OS1. In Fig. 2, one combination pair is shown, but corresponding tissue cells and organoids can be set as one combination pair, or 16 combination pairs.

[0079] M1 can calculate a first correlation coefficient between the expression level of the first gene in the tissue cells of the combination pair (CP1) and the expression level of the first gene in the organoids of the combination pair (CP1). The first correlation coefficient can be calculated using the relationship between gene expression levels, as illustrated in FIG. 5.

[0080] Tissue cells belonging to TS1 and organoids (OS1) corresponding to the tissue cells are each generated as one combination pair, and through this process, combination pairs can be generated as many times as the number of tissue cells (as many times as the number of organoids).

[0081] M2 can generate random correlation coefficients using the first method. M2 can randomly generate combination pairs (RC11, RC12) between tissue cells from TS1 and organoids from OS1. M2 can also generate sets of non-overlapping combination pairs, such as RS11, RS12, and RS13.

[0082] M2 can calculate correlation coefficients by comparing the expression levels of the first gene for tissue cells and organoids of random combination pairs in RS11, RS12, RS13, …, which are sets of combination pairs.

[0083] The correlation coefficients produced by M2 are referred to as the first random correlation coefficients. M2 can generate a first correlation coefficient distribution that represents the first random correlation coefficients produced from each random combination pair in order of magnitude. M2 can calculate the first probability value of the first correlation coefficient from the first correlation coefficient distribution.

[0084] M3 can search for a set of tissues (TS2) in which a first gene is expressed and a set of organoids (OS2) in which the first gene is expressed. M3 can generate a second combination pair (CP2) comprising tissue cells from TS2 and organoids from OS2. While Figure 2 illustrates one combination pair, corresponding tissue cells and organoids can be configured as one combination pair, or as many as 16 combination pairs.

[0085] M3 can calculate a second correlation coefficient based on the relationship between the expression level of the first gene in the tissue cells of the combination pair (CP2) and the expression level of the first gene in the organoids of the second combination pair (CP2). The second correlation coefficient can be calculated based on the relationship between the gene expression levels of 16 combination pairs. The method for calculating the second correlation coefficient is different from that for calculating the first correlation coefficient.

[0086] Tissue cells belonging to TS2 and organoids (OS2) corresponding to the tissue cells are each generated as a combination pair, and through this process, combination pairs can be generated as many times as the number of tissue cells (as many times as the number of organoids).

[0087] M4 can generate random correlation coefficients using the second method. Random combination pairs (RC21, RC22) can be generated by randomly generating combination pairs between tissue cells from TS2 and organoids from OS2. M4 can also generate sets of non-overlapping combination pairs, such as RS21, RS22, and RS23.

[0088] M4 can calculate correlation coefficients based on the relationship between the expression levels of the first gene of tissue cells and organoids in random combination pairs in a set of combination pairs, RS21, RS22, RS23, … According to embodiments of the present disclosure, the expression levels can be expressed as two-dimensional coordinates, and the overall relationship between the coordinates can be calculated as a correlation coefficient.

[0089] The correlation coefficients produced by M4 are referred to as second random correlation coefficients. M4 can generate a second correlation coefficient distribution that arranges the second random correlation coefficients produced from each random combination in order of magnitude. M4 can then produce a second probability value for the second correlation coefficient from the first correlation coefficient distribution.

[0090] M5 can produce organoids with a high level of tissue cell reproducibility for the first gene using the first probability value and the second probability value.

[0091] FIG. 3 is a drawing of a learning device (200) that calculates a threshold value for a reproduction level in a genetic reproducibility calculation device (100).

[0092] The genetic reproducibility calculation device (100) can calculate a reproduction level value for a plurality of organoids for an input gene. The genetic reproducibility calculation device (100) can calculate a threshold value determined based on the reproduction level values ​​for a plurality of organoids for a plurality of genes. The threshold value may be a threshold value for determining the presence or absence of a gene reproduction level of an organoid. For example, an organoid having a reproduction level value greater than or equal to the threshold value may be determined to have a gene reproduction similar to a tissue cell. An organoid having a reproduction level value less than the threshold value may be determined to have a gene reproduction not similar to a tissue cell. The genetic reproducibility calculation device (100) can input training data including genetic information and probability values ​​of correlation coefficients of a plurality of organoids into the learning device (200). The genetic reproducibility calculation device (100) can be trained with training data having as input combination pairs including a set of tissue cells and a set of organoids for each gene, and having as output a correlation coefficient which is a relationship between the combination pairs. Training data may include correlation coefficient values ​​between each gene's information and the tissue cells and organoids for each gene. These correlation coefficients can be considered a numerical indicator of how similar the gene expression patterns are between tissues and organoids.

[0093] The learning device (200) can be trained with training data using machine learning and / or deep learning. The learning device (200) can calculate a threshold for a correlation coefficient or probability value, which is a reproduction level of an organoid, using machine learning and / or deep learning. The learning device (200) can determine a threshold value that serves as a criterion for determining an organoid with a high reproduction level for tissue cells and a high reproduction level for genes. The learning device (200) can determine a threshold value that serves as a criterion as a result of learning by continuously inputting training data.

[0094] The learning device (200) can transmit the threshold value calculated as a learning result to the genetic reproducibility calculation device (100) to determine whether the organoid is reproducible or whether the reproducibility level is high or low.

[0095] Figure 4 is a table of the first and second correlation coefficients for each gene derived from the patient's lung cancer tissue and organoids derived from the lung cancer tissue.

[0096] The patient's lung cancer tissue may be obtained from lung cancer-related samples within the GBCC's patient-derived bank (PDO). However, the patient's lung cancer tissue may be obtained through various methods.

[0097]

[0098] *Forty-eight organoids can be derived from a patient's lung cancer tissue. Information about the organoid cells derived from each tissue can be stored in advance. Information about the organoids can be experimentally acquired and updated.

[0099] According to embodiments of the present disclosure, the genetic reproducibility calculation device can generate 48 pairs of combinations between lung cancer tissues and organoids derived therefrom. Each of the 48 pairs of combinations can be expressed by coordinates.

[0100] As illustrated in Figure 4, the relationship between gene expression levels for each gene between lung cancer tissue and organoids can be calculated. The relationships between gene expression levels can be tabulated by calculating Pearson correlation coefficients and Spearman correlation coefficients.

[0101] Figure 5 is a diagram showing the coordinates of combination pairs between lung cancer tissues and organoids.

[0102] As illustrated in Figure 5, a combination pair can be represented as a single coordinate. The expression levels in tissue cells and organoids included in the combination pair can be expressed as two-dimensional coordinates. The relationship between the coordinates derived for a single gene can be calculated, yielding a correlation coefficient of 0.8477.

[0103] Figure 6 is an exemplary distribution diagram illustrating first random correlation coefficients of random pairs between lung cancer tissue of a patient and organoids derived from lung cancer tissue.

[0104] For the lung cancer tissue of the patient obtained in FIG. 4 and the 48 organoids derived from the lung cancer tissue, the process of creating random combination pairs is repeated 1000 times, and the random correlation coefficients of the random combination pairs can be calculated as Pearson correlation coefficients and expressed as in FIG. 6. The average value of the first random correlation coefficients is most distributed at -0.0008258, and the standard deviation value is 0.20836312. The distribution of the first random correlation coefficient can be expressed as having an average value and a standard deviation value. In the first random correlation coefficients, with respect to the transformation of the probability value of the first correlation coefficient, the probability value of occurrence in the range between -standard deviation and +standard deviation based on the mean value has 68.27 percent, the probability value of occurrence in the range between -2 times the standard deviation and +2 times the standard deviation based on the mean value has 95.45 percent, and the probability value of occurrence in the range between -3 times the standard deviation and +3 times the standard deviation based on the mean value has 97.73 percent.

[0105] The position of the correlation coefficient is calculated using the mean and standard deviation in the distribution of random correlation coefficients, and the reproduction level of the organoid is calculated based on the probability value of the correlation coefficient from the position value of the correlation coefficient.

[0106] Figure 7 is an exemplary distribution diagram illustrating second random correlation coefficients of random pairs between lung cancer tissue of a patient and organoids derived from lung cancer tissue.

[0107] For the lung cancer tissue of the patient obtained in Fig. 4 and the 48 organoids derived from the lung cancer tissue, the process of creating random combination pairs is repeated 1000 times, and the random correlation coefficients of the random combination pairs are calculated as the spearman correlation coefficient and expressed as in Fig. 6.

[0108] The mean of the second random correlation coefficients is most commonly distributed at -0.00081206, and the standard deviation is 0.20838651. The distribution of the second random correlation coefficients can be expressed as having a mean value and a standard deviation value.

[0109] The position of the second correlation coefficient is calculated using the mean and standard deviation in the distribution of the second random correlation coefficients, and the reproduction level of the organoid is calculated based on the probability value of the correlation coefficient from the position value of the second correlation coefficient.

[0110] Figure 8 is a flowchart of a method for calculating genetic reproducibility according to embodiments of the present disclosure.

[0111] In S110, the genetic reproducibility calculation device (100) can input information about the first gene into the data input unit (110). The genetic reproducibility calculation device (100) can search for tissue cells (hereinafter, a set of tissues) in which the first gene is expressed, and organoids (hereinafter, a set of organoids) corresponding to the tissue cells.

[0112] A genetic reproducibility calculation device (100) can generate combination pairs (hereinafter, a set of combination pairs) between one or more tissue cells included in a set of tissues and one or more organoids included in a set of organoids. The set of tissues may be one tissue, and a plurality of organoids derived from one tissue may be included in the set of organoids, but is not limited thereto.

[0113] A single combination pair included in a set of combination pairs is referred to as a first combination pair. The first combination pair may include a first tissue cell and a first organoid that correspond to each other. The first tissue cell may be included in a set of tissues, and the first organoid may be included in a set of organoids.

[0114] In S120, the genetic reproducibility calculation device (100) can calculate a first correlation coefficient between the gene expression level of the first tissue cell included in the first combination pair and the gene expression level of the first organoid. Here, the gene expression level can be measured based on RNA sequencing data for the organoid or tissue cell. Furthermore, the first correlation coefficient can be viewed as an indicator of how similar the gene expression patterns are between the tissue and the organoid.

[0115] The first correlation coefficient can be calculated by the mathematical formula below.

[0116]

[0117] Here, X may be the expression level of the first gene of the tissue cell, and Y may be the expression level of the first gene of the organoid.

[0118] In S130, the genetic reproducibility calculation device (100) can calculate a second correlation coefficient between the gene expression amount of the first tissue cell included in the first combination pair and the gene expression amount of the first organoid. The second correlation coefficient may be calculated by a correlation coefficient method different from the first correlation coefficient. For example, the first correlation coefficient may be Pearson's correlation coefficient, and the second correlation coefficient may be Spearman's correlation coefficient, but is not limited thereto, and may be at least one of Kendall's Tau, Distance Correlation, Hoeffiding's D, and Maximal Information Coefficient. In addition, the second correlation coefficient may be viewed as an indicator that numerically represents how similar the gene expression patterns in the tissue and the organoid are.

[0119] In S140, the genetic reproducibility calculation device (100) can generate n random combination pairs by randomly matching one or more tissue cells included in a set of tissue cells and one or more organoids included in a set of organoids in a one-to-one relationship. Through the above process, the genetic reproducibility calculation device (100) can generate first random combination pairs, second random combination pairs, third random combination pairs, … n-th random combination pairs. The first to n-th random combination pairs can be generated without overlapping each other. n is a natural number, and can refer to the number of methods for generating random combination pairs without overlapping between a set of tissue cells and a set of organoids.

[0120] The genetic reproducibility calculation device (100) can calculate correlation coefficients between the gene expression levels of tissue cells of the first random combination pairs and the gene expression levels of the organoid. In this manner, the correlation coefficient calculation unit (120) can calculate correlation coefficients between the gene expression levels of tissue cells of the second to n-th random combination pairs and the gene expression levels of the organoid. The first correlation coefficient calculation unit (120) can set correlation coefficients calculated by the first correlation coefficient method for the first to n-th random combination pairs as first random correlation coefficients. The first correlation coefficient method can be at least one of Pearson, Spearman, Kendall's Tau, Distance Correlation, Hoeffiding's D, and Maximal Information Coefficient. The genetic reproducibility calculation device (100) can generate a first correlation coefficient distribution by arranging the first random correlation coefficients in order of size. The genetic reproducibility calculation device (100) can generate a probability value of the first correlation coefficient in the first correlation coefficient distribution. More specifically, the probability value of the first correlation coefficient may be the statistical significance value (P) of the first correlation coefficient with respect to the first random correlation coefficients. The statistical significance value may be calculated in various ways.

[0121] In S150, the genetic reproducibility calculation device (100) can generate n random combination pairs by randomly matching one or more tissue cells included in a set of tissue cells and one or more organoids included in a set of organoids in a one-to-one relationship. Through the above process, the genetic reproducibility calculation device (100) can generate first random combination pairs, second random combination pairs, third random combination pairs, … n-th random combination pairs. The first to n-th random combination pairs can be generated without overlapping each other.

[0122] The genetic reproducibility calculation device (100) can calculate correlation coefficients between the gene expression levels of tissue cells of the first random combination pairs and the gene expression levels of the organoid. In this way, the genetic reproducibility calculation device (100) can calculate correlation coefficients between the gene expression levels of tissue cells of the second to n-th random combination pairs and the gene expression levels of the organoid. The genetic reproducibility calculation device (100) can set correlation coefficients calculated by the second correlation coefficient method for the first to n-th random combination pairs as second random correlation coefficients. The second correlation coefficient method is different from the first correlation coefficient method and can be at least one of Pearson, Spearman, Kendall's Tau, Distance Correlation, Hoeffiding's D, and Maximal Information Coefficient. The genetic reproducibility calculation device (100) can generate a second correlation coefficient distribution map by considering the second random correlation coefficients. The genetic reproducibility calculation device (100) can generate a probability value of the second correlation coefficient in the second correlation coefficient distribution map. More specifically, the probability value of the second correlation coefficient may be the statistical significance value (P) of the second correlation coefficient.

[0123] In S150, the gene reproducibility calculation device (100) can calculate the reproduction level of the first gene in the organoid by considering the first probability value corresponding to the position value of the first correlation coefficient and the second probability value corresponding to the position value of the second correlation coefficient. At this time, the gene reproducibility calculation device (100) can calculate the statistical significance value for the first probability value and the second probability value using a method for calculating statistical significance, for example, the Stouffer method. The reproduction level of the first gene in the organoid can be calculated based on the statistical significance value calculated in this way. The reproduction level can be determined to be higher as the first and / or second probability values ​​are larger.

[0124] Additionally, the gene reproducibility calculation device (100) can generate output data regarding the level of reproducibility of the first gene in the organoid. The output data can include at least one of a distribution of the first and / or second correlation coefficients, the first and / or second probability values, and the first and / or correlation coefficients. The output data can include data visually representing the level of reproducibility of the first gene in the organoid. The output data can include the presence or absence of reproducibility (Yes / No) or whether the level of reproducibility is high or low in comparison with a predetermined threshold.

[0125] In addition, although the preferred embodiments of the present invention have been illustrated and described above, the present invention is not limited to the specific embodiments described above, and various modifications can be made by a person having ordinary skill in the art to which the present invention pertains without departing from the gist of the present invention claimed in the claims, and such modifications should not be understood individually from the technical idea or prospect of the present invention.

[0126] Therefore, the spirit of the present invention should not be limited to the embodiments described above, and not only the scope of the patent claims described below, but also all scopes equivalent to or equivalently modified from the scope of the patent claims are considered to fall within the scope of the spirit of the present invention.

Claims

1. A step of a genetic reproducibility calculation device receiving information about a first gene, searching for first tissue cells expressing the first gene and first organoids corresponding to the first tissue cells, and setting a combination pair including one of the first tissue cells and one of the first organoids; A step in which the genetic reproducibility calculation device calculates a first correlation coefficient between the expression levels of the first gene for each combination pair using a first method; A step in which the genetic reproducibility calculation device calculates a second correlation coefficient between the expression levels of the first gene for each combination pair using a second method different from the first method; The step of the genetic reproducibility calculation device setting random combination pairs for the first tissue cells and the first organoids, calculating first random correlation coefficients between the expression levels of the first gene for the random combination pairs using a first method, and calculating a probability value of the first correlation coefficient in a distribution of the first random correlation coefficients; A step of the genetic reproducibility calculation device calculating second random correlation coefficients between the expression levels of the first gene for the random combination pairs using a second method, and calculating a probability value of the second correlation coefficient in a distribution of the second random correlation coefficients; A method for calculating gene reproducibility, comprising a step of calculating a reproducibility level of a first gene based on a probability value of the first correlation coefficient or a probability value of the second correlation coefficient, by the gene reproducibility calculation device.

2. In paragraph 1, A method for calculating genetic reproducibility, wherein the first correlation coefficient calculated by the first method is a Pearson correlation coefficient.

3. In paragraph 1, A method for calculating genetic reproducibility, wherein the second correlation coefficient calculated by the second method is a Spearman correlation coefficient.

4. In paragraph 1, A method for calculating genetic reproducibility, wherein the genetic reproducibility calculation device further comprises a step of generating data on the reproducibility level as output data and transmitting the data to another device.

5. In paragraph 1, The step of calculating the reproduction level of the first gene is A method for calculating gene reproducibility, wherein the reproducibility level of the first gene is determined as present or absent based on whether the probability value of the first correlation coefficient is greater than or equal to a first threshold value, or whether the probability value of the second correlation coefficient is greater than or equal to a second threshold value.

6. In paragraph 5, The first or second threshold value is, A method for calculating genetic reproducibility, which is a value output from a learning device that learns using machine learning or deep learning.

7. In paragraph 6, The above learning device, A method for calculating gene reproducibility, which is learned by inputting genetic information, correlation coefficients calculated by comparing gene expression levels, or probability values ​​of correlation coefficients.

8. A data input unit that inputs information about a first gene, searches for first tissue cells in which the first gene is expressed, and first organoids corresponding to the first tissue cells, and sets a combination pair including one of the first tissue cells and one of the first organoids; A first correlation coefficient calculation unit for calculating a first correlation coefficient between the expression levels of the first gene for each combination pair using a first method; A second method different from the first method includes a second correlation coefficient calculation unit that calculates a second correlation coefficient between the expression levels of the first gene for each combination pair using the second method; The first correlation coefficient calculation unit sets random combination pairs for the first tissue cells and the first organoids, calculates first random correlation coefficients between the expression levels of the first gene for the random combination pairs using a first method, and calculates the probability value of the first correlation coefficient in the distribution of the first random correlation coefficients. The second correlation coefficient calculation unit calculates second random correlation coefficients between the expression levels of the first gene for the random combination pairs using a second method, and calculates the probability value of the second correlation coefficient in the distribution of the second random correlation coefficients. A genetic reproducibility calculation device further comprising a reproducibility level calculation unit that calculates a reproducibility level of a first gene based on a probability value of the first correlation coefficient or a probability value of the second correlation coefficient.

9. In paragraph 8, A genetic reproducibility calculation device in which the first correlation coefficient calculated by the first method is a Pearson correlation coefficient.

10. In paragraph 8, A genetic reproducibility calculation device, wherein the second correlation coefficient calculated by the second method is a Spearman correlation coefficient.

11. In paragraph 8, A genetic reproducibility calculation device further comprising an output data generation unit that generates data for the above reproducibility level as output data and transmits it to another device.

12. In paragraph 8, The above reproduction level calculation unit is, A genetic reproducibility calculation device that determines the presence or absence of the reproducibility level of the first gene based on whether the probability value of the first correlation coefficient is greater than or equal to a first threshold value or whether the probability value of the second correlation coefficient is greater than or equal to a second threshold value.

13. In paragraph 12, The first or second threshold value is, A genetic reproducibility calculation device, which is a value output from a learning device that learns using machine learning or deep learning.

14. In paragraph 13, The above learning device, A genetic reproducibility calculation device that learns by inputting genetic information, correlation coefficients calculated by comparing gene expression levels, and probability values ​​of correlation coefficients.

15. A computer program stored in a computer-readable storage medium to execute the method of paragraph 1 using a computer.

Images