Morphological analysis device, morphological analysis method, and morphological analysis program

Abstract

Provided are a morphology analysis device, a morphology analysis method, and a morphology analysis program with which the morphology of a constituent element of a molecule can be recognized. A morphology analysis device (200) for analyzing the morphology of a molecule in a solution is provided with: a reference data storage unit (234) for storing reference data, which is molecular shape data of a predetermined test molecule; a target data storage unit (235) that stores target data, which is molecular shape data of the molecule to be analyzed; and a morphology estimation unit (237) that estimates the morphology of a constituent element of a molecule in the subject data by specifying the state of a physical quantity constituting the molecule shape of the subject data with respect to the reference data, both the reference data and the subject data having a molecular-level resolution.

Description

Technical Field

[0001] The present invention relates to a morphology analysis device, a morphology analysis method and a morphology analysis program for analyzing the morphology of molecules in a solution. Background Art

[0002] Biopharmaceuticals are proteins, containing macromolecules with large molecular weight and complex structures as main components. Unlike low molecular weight pharmaceuticals, biopharmaceuticals cannot produce chemically uniform proteins, so conformation, heterogeneity of sugar chain structure, non-specific association, etc. have a great impact on their effectiveness or safety. Therefore, in order to ensure reliability, post-translational modification, aggregation analysis, and analysis of three-dimensional structural changes are required. The manufacture of biopharmaceuticals requires large-scale equipment investment, and quality control is very difficult (Non-Patent Literature 1).

[0003] When administering biopharmaceuticals, not only is the dosage large, but thousands of times the amount of drug substance is required for a single administration. In addition, the structure and quality of the drug substance as the final product must be evaluated, and the comparability of products including various parameters must be ensured.

[0004] Even if biopharmaceuticals are produced using the same gene sequence, the conformations, sugar chain structures, or associations of the progenitors are not the same pharmaceutical products. Therefore, in order to ensure the quality within certain standards, the conditions of culture, purification, or concentration must be strictly managed.

[0005] In particular, in the production of antibodies using CHO cell lines, long-term culture will reduce the antibody production ability and discharge impurities. Therefore, it is necessary to always continue to culture fresh cells. In addition, it is difficult to control the expression regulation, and sometimes poor quality antibodies are produced. Therefore, for comprehensive quality management, it is necessary not only to analyze the conventional chemical properties and comprehensive physicochemical properties, but also to confirm structural changes or fluctuations.

[0006] However, the existing analysis or analysis process of biopharmaceuticals takes a long time, and before analysis or analysis, it is necessary to develop in advance an analysis or analysis method corresponding to the state of the sample. Moreover, in the three-dimensional structure analysis technology, which is an extremely limited method, there is a decisive restriction that it cannot be measured in the original state of the solution. Therefore, for the characteristics related to the three-dimensional structure, it is impossible to measure or analyze the characteristics of the original body under the target conditions or state. In other words, not only is it impossible to know the conformation of the original body molecule at that time point, it is also impossible to know the distribution of mobility, etc. In this way, it has not yet been possible to implement morphological analysis of molecules that is effective in process management for the development or manufacture of pharmaceuticals.

[0007] In the past, there were also methods such as cryo-electron microscopy that freeze the sample to observe biopolymers, but since the sample is frozen, the mobility of the biopolymer will not be revealed. In addition, it is also known to use a database to obtain a three-dimensional distribution of electron density based on the coordinates of atoms constituting proteins with known structures, but the movable structure cannot be identified based on the obtained three-dimensional distribution (Patent Document 1). Moreover, in either case, it is obvious that the observation results under given conditions required in the process of formulation development or quality management cannot be obtained.

[0008] Prior art literature

[0009] Non-patent literature

[0010] Non-patent document 1: "Current status and issues of the biopharmaceutical industry", Junichi Matsuzaki, Journal of the Society of Biotechnology, 91(9), 495-498, 2013

[0011] Patent Literature

[0012] Patent Document 1: Japanese Patent Application Publication No. 2005-250721 Summary of the invention

[0013] Problem that the invention aims to solve

[0014] As described above, it is required to be able to apply to the observation technology of three-dimensional structure correlation information of quality management headed by the formulation development or manufacturing process of biopharmaceuticals. However, in the prior art, it is impossible to determine the electron density of the constituent elements that reflect the conformational changes or mobility of the biopharmaceutical molecules in the solution state. Therefore, it is impossible to fully analyze the molecular form of the formulation solution components. On the other hand, if there is a structural visualization technology that has advantages in the visualization of the shape of molecules in the solution and the analysis of the fluctuations of molecules, it is possible to evaluate the conformational changes or mobility in the solution of biopharmaceutical molecules actually existing in proteins and the like.

[0015] The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a morphological analysis device, a morphological analysis method, and a morphological analysis program capable of identifying the morphology of constituent elements of a molecule.

[0016] Solutions for solving problems

[0017] (1) In order to achieve the above-mentioned purpose, the morphological analysis device of the present invention is a morphological analysis device that analyzes the morphology of molecules in a solution, and is characterized by comprising: a reference data storage unit that stores reference data, wherein the reference data is molecular shape data of the detected molecule determined in advance; an object data storage unit that stores object data, wherein the object data is molecular shape data of the detected molecule that is the object of analysis; and a morphological estimation unit that estimates the morphology of constituent elements of the molecules in the object data by determining the state of physical quantities of the constituent molecular shapes of the object data relative to the reference data, wherein both the reference data and the object data have a resolution at the molecular level.

[0018] (2) In addition, in the morphological analysis device described in (1) above, the morphology of the constituent element is a physical difference directly derived from a chemical difference between the constituent element of the molecule specified by the object data and the molecule specified by the reference data.

[0019] (3) In addition, in the morphological analysis device described in (1) above, the morphology of the constituent element is a difference in at least one of conformation and motility between the molecule specified by the object data and the molecule specified by the reference data.

[0020] (4) In addition, in the morphological analysis device described in (2) or (3) above, the molecule specified by the target data is originally different from the molecule specified by the reference data.

[0021] (5) In addition, in the morphological analysis device described in (2) or (3) above, the molecule specified by the target data is the same substance as the molecule specified by the reference data at a different time point.

[0022] (6) In addition, in the morphological analysis device described in any one of (1) to (5) above, it is characterized in that the morphological estimation unit determines the state of the physical quantities constituting the molecular shape by the height of the physical quantities constituting the molecular shape, or the extended shape or size of the physical quantities as their aggregate.

[0023] (7) In addition, in the morphological analysis device described in any one of (1) to (6) above, the morphological estimation unit uses data of physical quantities constituting the molecular shape obtained by measuring a standard sample with low motility as the reference data.

[0024] (8) In the morphological analysis device described in any one of (1) to (6) above, the morphological estimation unit uses the accumulated analysis data as the reference data for the teacher data.

[0025] (9) In addition, in the morphological analysis device described in any one of (1) to (8) above, it is characterized in that it also includes an output information generating unit that generates output information in a manner that can identify the morphology of the object data, the output information generating unit displays the object data on an output device, and accepts input of designation information that specifies a constituent element whose morphology is to be inferred for the displayed object data, and the morphology inference unit infers the morphology of the constituent element specified based on the designation information.

[0026] (10) In addition, in the morphological analysis device described in any one of (1) to (9) above, it is characterized in that it also has an output information generating unit that generates output information in a manner capable of identifying the morphology of the object data, and the output information generating unit displays the object data on an output device and displays it by superimposing a specific mark representing the estimated morphology on the object data.

[0027] (11) In addition, in the morphological analysis device described in any one of (1) to (10) above, the test molecule is a Y-shaped antibody molecule, and the constituent element is Fab or Fc.

[0028] (12) In addition, the morphological analysis device described in (11) above is characterized in that it also comprises: an output information generating unit, which generates output information in a manner capable of identifying the morphology of the object data; and a quality determination unit, which determines whether the detected molecule meets the quality standard based on the estimated morphology, and when the detected molecule does not meet the quality standard, the output information generating unit displays the detected molecule as not meeting the quality standard.

[0029] (13) In addition, the morphological analysis method of the present invention is a morphological analysis method for analyzing the morphology of molecules in a solution, and is characterized in that it includes: a step of reading reference data, wherein the reference data is molecular shape data of the molecule to be tested that is determined in advance; a step of reading object data, wherein the object data is molecular shape data of the molecule to be tested that is the object of analysis; and a step of estimating the morphology of constituent elements of the molecules in the object data by determining the state of physical quantities of the molecular shapes constituting the object data relative to the reference data, wherein both the reference data and the object data have a resolution at the molecular level.

[0030] (14) In addition, the morphological analysis program of the present invention is a morphological analysis program for analyzing the morphology of molecules in a solution, and is characterized in that a computer is caused to execute: a process of reading reference data, wherein the reference data is molecular shape data of a molecule to be tested that is determined in advance; a process of reading object data, wherein the object data is molecular shape data of the molecule to be tested that is an object of analysis; and a process of estimating the morphology of constituent elements of the molecules in the object data by determining the state of physical quantities of the molecular shapes constituting the object data relative to the reference data, wherein both the reference data and the object data have a resolution at the molecular level. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a schematic diagram showing the morphological analysis system of the present invention.

[0032] Figure 2 It is a perspective view showing a small-angle X-ray scattering device.

[0033] Figure 3 It is a block diagram showing the details of the morphological analysis unit.

[0034] Figure 4 is a flowchart showing a morphological analysis method.

[0035] Figure 5 (a) and (b) are diagrams showing a projection image of the electron density of target data and identified components, respectively.

[0036] Figure 6 (a) and (b) are diagrams showing marks (arrows), respectively.

[0037] Figure 7 (a) to (c) are diagrams showing projection images of electron density and marks (arrows) when antibody molecules to be analyzed are viewed from directions perpendicular to each other.

[0038] Figure 8 (a) and (b) are diagrams showing marks (encircling lines), respectively.

[0039] Fig. 9 This is a schematic diagram showing the production process of biopharmaceuticals.

[0040] Fig.10 is a block diagram showing details of a structure determination unit.

[0041] Fig.11 is a flow chart illustrating a structure determination method.

[0042] Fig.12 It is a schematic diagram showing a structural model.

[0043] Fig.13 (a) and (b) show the ideal distribution and the unanalyzable χ 2 - Coordinate diagram of the Rg(ind) distribution.

[0044] Fig.14 (a) and (b) are diagrams showing projection images of the electron density of a starting antibody molecule and an antibody-drug complex, respectively.

[0045] Fig.15 (a) and (b) are diagrams showing projection images of the electron density of antibody molecules one day after production and two months after production, respectively. Fig.15 (b) is a mark showing the opening degree of Fab.

[0046] Fig.16 The structure models and their Rg(ind)-χ 2 Schematic representation of points on a coordinate graph.

[0047] Fig.17 The Fab lengths of the left and right Fabs of the Y-shaped antibody molecules are shown in Figure 1. 2 Table of calculation examples for corresponding weightings. DETAILED DESCRIPTION

[0048] Next, embodiments of the present invention will be described with reference to the accompanying drawings. In order to facilitate understanding of the description, the same components are denoted by the same reference numerals in the drawings, and repeated descriptions are omitted.

[0049] In addition, below, as a representative of the physical quantities constituting the molecular shape, the present invention is described using electron density, but the electron density distribution can also be replaced with a physical quantity constituting the molecular shape such as a potential map, and the electron density data can also be replaced with molecular shape data as data of the physical quantity constituting the molecular shape. The physical quantity constituting the molecular shape refers to quantitative information corresponding to the position that can determine the structure of the molecule through interaction with radiation. The physical quantity constituting the molecular shape includes, for example, an electron density map, or a potential map as a three-dimensional map representing the potential within the molecule. Molecular shape data refers not only to the case of a single structural model, but sometimes also refers to multiple structural models with distribution, as described later.

[0050] [principle]

[0051] When X-rays are irradiated on molecules moving freely in a solution, scattered light is generated in the form of rings rather than points. By detecting such scattered light and selecting an electron density map obtained by recursive phase improvement, a structural model obtained by fitting an X-ray scattering profile calculated using an assumed coordinate structural model, or a structural model with a high fit consistency index among a plurality of structural models constituting their ensemble, the structural model of molecules in a solution that adopt a structure with dynamic fluctuations can be accurately reproduced.

[0052] When using such a structural analysis method, the distribution of the electron density of the molecules in the solution can be determined. The constituent elements of the molecule are sometimes at an angle to each other as in the benchmark, and sometimes the two ends are opened to more than the benchmark. If these situations are compared, it can be said that there are differences in the mutual conformations. In addition, there are also cases where only one of the constituent elements is fixed and the other constituent element can swing in a certain direction. In such a case, it can be said that the constituent element is mobile. In the molecules determined by the structural model, the electron density of the constituent element with mobility is locally reduced, and its distribution has an extension in the entire range of motion. In this way, by using the distribution of electron density, the morphology of the constituent elements of the molecule can be estimated. The morphology of the constituent elements of the molecule estimated not only includes pure physical differences that are independent of the chemical changes between the comparison objects, physical differences indirectly caused by chemical differences, but also includes physical differences directly derived from chemical differences. The details of the morphology of the constituent elements of the molecule will be described later.

[0053] [Morphological analysis system]

[0054] Figure 1is a schematic diagram showing a morphological analysis system 10. The morphological analysis system 10 includes a small-angle X-ray scattering device 100 and a morphological analysis device 200. The small-angle X-ray scattering device 100 measures the small-angle X-ray scattering distribution diagram by irradiating the sample S0 with X-rays and detecting the small-angle scattered X-rays. The sample S0 is suitable for polymers in solution, especially biopolymers. The sample S0 is preferably a molecule, molecular complex or structure for medical use in solution. By using the X-ray solution scattering method by the morphological analysis system 10, the combination (overall) of the structure of the biopolymer that cannot be observed in a frozen or crystallized state can be visualized.

[0055] The morphology analysis device 200 is composed of a computer 205 , an input device 280 , and an output device 290 , and controls the operation of the small-angle X-ray scattering device 100 , and acquires and processes measurement data from the small-angle X-ray scattering device 100 .

[0056] The X-ray small angle scattering device 100 includes an X-ray generating unit 110, a sample loading mechanism 120, a detector 130, and a control unit 140. The X-ray generating unit 110 has an X-ray source 111, and irradiates the sample S0 with X-rays. The sample loading mechanism 120 sends a polymer in a solution as a sample or a solvent without a polymer to an X-ray irradiation position. The detector 130 detects the X-rays scattered by the sample S0, and sends the obtained measurement data to the computer 205.

[0057] The computer 205 is, for example, a PC, and is composed of a processor that executes processing and a memory or hard disk that stores programs or data. The computer 205 receives input from a user from an input device 280 such as a keyboard and a mouse. On the other hand, the computer 205 displays a drawing (plot; plotting), a visualized polymer image, an input screen, etc. on an output device 290 such as a display. The computer 205 can be a server device placed on the cloud. In addition, from the perspective of processing burden, the function of controlling the operation of the small-angle X-ray scattering device 100 and the function of processing the measurement data can also be separated, and the control can be performed by a PC installed on site, and the data processing can be performed by the server device.

[0058] [X-ray small angle scattering device]

[0059] Figure 21 is a perspective view showing a small-angle X-ray scattering device 100. The small-angle X-ray scattering device 100 has an X-ray source 111, an optical system 115, a Kratzky block 117, a sample holding tube 125, and a detector 130. The X-ray source 111 is a linear radiation source or a point radiation source, and releases a scattered beam. The optical system 115 is, for example, a KB parallel or serial optical system. A pair of Kratzky blocks 117 define one side and the other side of the X-ray beam by interacting with the X-rays at their respective edges. In this way, parasitic scattering can be removed from the irradiated X-rays. The sample holding tube 125 delivers and holds 5μl to 10μl of a solution sample. The detector 130 detects X-rays scattered by the solution sample.

[0060] [Morphological analysis device]

[0061] Figure 3 2 is a block diagram showing the details of the morphological analysis unit. The morphological analysis device 200 analyzes the morphology of the polymer in the solution whose structure has been determined. The functions of the morphological analysis device 200 are mainly realized by the computer 205. The computer 205 has a basic function unit 210, a structure determination unit 220 and a morphological analysis unit 230. The various parts constituting each unit can send and receive information to each other through the control bus L.

[0062] (Basic functional unit)

[0063] The basic function unit 210 performs basic functions such as input / output to the user and measurement control. The basic function unit 210 includes an input / output control unit 211, a measurement control unit 215, and a measurement data storage unit 217. In addition, each storage unit in the morphological analysis device 200, such as the measurement data storage unit 217, may be divided into multiple units only in terms of function, and may be physically equivalent to the same memory or any one of multiple memories.

[0064] The input / output control unit 211 controls input / output to the outside. Specifically, it accepts input from the input device 280 and controls output to the output device 290. The input / output control unit 211 can accept input of measurement conditions or input of conditions for generating a plurality of structural models, for example. In addition, it can output various drawings or output the determined polymer structure.

[0065] The input / output control unit 211 displays information generated for output on the output device 2900. The information generated for output may include the electron density of the test molecule obtained by measurement. The input / output control unit 211 receives input of designation information for designating a component of a desired form of analysis for the displayed object data.

[0066] The measurement control unit 215 controls the operation of the small-angle X-ray scattering device 100. Examples of the control include the sending of samples, the generation of X-rays, and the movement of the sample position and the detector. The control instructions are sent to the control unit 140 in the small-angle X-ray scattering device 100, thereby controlling each part of the small-angle X-ray scattering device 100.

[0067] The measurement data storage unit 217 stores measurement data of the X-ray small-angle scattering distribution pattern detected by the X-ray small-angle scattering device 100. The stored measurement data is used for generating a structural model, calculating an index, or calculating an actual value of the molecular size of a polymer.

[0068] (Structure determination unit)

[0069] The structure determination unit 220 generates a plurality of structural models based on the measured X-ray scattering distribution map obtained by measuring the sample. Then, the consistency between the calculated X-ray scattering distribution map calculated based on each of the plurality of structural models and the measured X-ray scattering distribution map is calculated, and a representative structural model is selected from the plurality of structural models based on the consistency. In this way, the structure determination unit 220 accurately reproduces the structural model of the polymer in the solution that adopts a structure with dynamic fluctuations. The details of the structure determination unit 220 will be described later.

[0070] (Morphological Analysis Unit)

[0071] The morphology analysis unit 230 includes an object data extraction unit 231, a reference data storage unit 234, an object data storage unit 235, an output information generation unit 236, a morphology estimation unit 237, and a quality determination unit 238. The morphology analysis unit 230 analyzes the morphology of molecules in the solution based on the electron density of the object data obtained by the structure determination unit 220.

[0072] The object data extraction unit 231 extracts object data of an object for analyzing the form of a constituent element from the structure model determined by the structure determination unit 220. The object data is generated by the structure determination unit 220 selecting a representative structure model from a plurality of structure models. The electron density data that may become the object data may be a representative structure model or a plurality of representative structure models with distribution. The distribution may be based on the x-values ​​in an objective evaluation without arbitrariness. 2 The values ​​of χ are weighted in multiple structural models, or they can be in the range of 2 =1 or more identical structural models within a specified range.

[0073] For example, it is possible to 2 The closer the value is to 1, the larger the value is defined, and each parameter is expressed as a weighted average.2 The function whose absolute value of difference from 1 decreases monotonically can use the following mathematical formula (1) or (2) as a weight function. In either case, a is a real number (usually 0), b and c are positive real numbers, and n is the base of the logarithm.

[0074] [Mathematical formula 1]

[0075] w=1 / 1+a+log n (x 2 )| b (1)

[0076] [Mathematical formula 2]

[0077] w=1 / (1+a+|χ 2 -1| b ) c (2)

[0078] In addition, in the above example, although χ 2 The reference value is 1, but the molecule to be analyzed takes two states, and when polydispersity occurs, the reference value may deviate from 1. In this case, the reference value that takes the deviation into account is calculated by 2 The weight is defined so that the closer to the reference value, the larger the value. For example, when analyzing PCT characteristics and two components coexist, an offset (a) may be required.

[0079] The reference data storage unit 234 stores reference data. Reference data refers to electron density data of a molecule to be tested that has been determined in advance. Reference data is, for example, registered as known electron density data of a molecule to be tested. The reference data is preferably analysis data previously analyzed and accumulated by the morphological analysis device 200, but may also be data consisting of known information. "In advance" means before the morphological analysis currently being performed. The electron density data that may serve as reference data may also be a representative structural model or a plurality of representative structural models with distribution, as in the case of the object data.

[0080] The object data storage unit 235 stores object data representing the electron density of the molecule to be analyzed. The object data refers to the electron density data of the molecule to be analyzed. Both the reference data and the object data have a molecular resolution. The "molecular resolution" is, for example, the resolution of the constituent elements of the molecule to be analyzed. Above and The following dimensions are preferred Above and The following resolution. In addition, the target data is data obtained by measurement, but the reference data does not necessarily have to be data obtained by measurement. The reference data may be data calculated based on a known crystal structure, for example.

[0081] The output information generating unit 236 generates output information in a manner that can identify the form of the object data. The output information generating unit 236 preferably generates output information based on the determined form of the molecule to be tested. The output information generating unit 236, for example, converts the electron density distribution projected in a specified direction into shades of a single color or shades of warm and cold colors corresponding to the position. For example, a position with a large electron density can be represented by a single dark color or a dark warm color. In addition, a position with a small electron density can be represented by a single light color or a light warm color. In addition, the distribution of electron density can also be represented by a gradient of two colors, such as a gradient from blue to red. Thus, for the constituent elements of the molecule to be tested, differences originating from at least one of the conformation and mobility can be represented.

[0082] Furthermore, the output information generating unit 236 may generate output information that visually displays the difference between the target data and the reference data, not based on the estimated morphology of the test molecule. For example, the output information may be generated by superimposing the target data on the reference data.

[0083] The output information generating unit 236 extracts the state or condition corresponding to the detected molecule and displays the state or condition together with the electron density distribution of the detected molecule. For example, the state or condition of the antibody molecule may include the presence or absence of sugar chains, changes over time, culture conditions, nutrient conditions, extraction methods, and transportation methods.

[0084] The output information generating unit 236 can also display the label together with the electron density of the detected molecule as the result of analyzing the morphology of the constituent elements. The display of the label is preferably superimposed on the display of the electron density of the constituent elements. In this way, the parameters of the morphology of the molecule can be displayed visually and easily. For example, the characteristics of the conformation can be visualized by the degree of opening of the Y-shape, or the direction or distance of the motility can be displayed by a line segment with arrows at both ends (arrow mark). For example, the propagation of the activity of the detected molecule can also be displayed by arrows.

[0085] In addition, a balloon-shaped encircling line (encircling line mark) can be displayed as a mark at a void site or a site with low density. It is effective to use a mark that can show the size of the space for the void, etc. The encircling line mark is an identifier showing the structure of the molecule, which indicates the physical difference of the constituent elements that induce the void, etc., which is directly derived from the chemical difference. In addition, in the change over time, the site of increased mobility can also be displayed by a mark.

[0086] Furthermore, when the molecule analyzed for morphology does not meet the quality standard, it can be displayed that the molecule does not meet the quality standard. As a case where the quality standard is not met, a parameter representing the morphology of a constituent element of a molecule is not within a predetermined value range.

[0087] The morphology estimation unit 237 estimates the morphology of the constituent elements of the molecule by reading the reference data and the object data and determining the state of the electron density of the object data relative to the reference data. Thus, as the morphology of the constituent elements of the molecule, a difference derived from at least one of conformation and mobility can be identified. As a result, for example, in the manufacturing process of biopharmaceuticals, the difference derived from at least one of conformation and mobility of the evaluated molecular structure can be used for quality management.

[0088] The morphology estimation unit 237 determines the state of the electron density distribution by the height, shape or size of the electron density. Thus, the morphology of the constituent elements of the molecule can be identified, and as a result, the morphology of the molecule under test can be estimated. In addition, as a size, for example, a cross-sectional area or a distance can be cited.

[0089] The morphology estimation unit 237 can, for example, determine each Fab or Fc of an antibody molecule as a constituent element of the molecule to be tested. For example, each constituent element can be determined by finding the center position of the molecule to be tested and finding the edge based on the center position. The constituent element can be regarded as an ellipsoid and the length of the short axis containing its center can be calculated as the size of the constituent element. The degree of opening can also be determined by determining the central axis of each constituent element and finding the angle between them. Such morphological information can also be calculated as a difference or ratio between the object data and the reference data.

[0090] Conformation is the spatial configuration of domains in a molecule. Spatial configuration is related to the positional information of electron density and includes the relative positions or angles of constituent elements. A change in conformation refers to a change from the spatial configuration of a domain at a certain measurement time to the spatial configuration of a domain at another measurement time. Motility is related to the level of electron density and is the activity of mobile constituent elements within a molecule. Motility refers to a state of being active without converging to a prescribed configuration within a certain measurement time range. Motility includes the speed, distance, range or direction of movement during movement.

[0091] The morphology estimation unit 237 can also evaluate the structural changes of the test molecule, determine the degree of change over time, evaluate the flexibility, or determine the soft part based on the difference in at least one of the conformation and mobility of the test molecule. For example, the difference in the degree of opening or flexibility of each Fab of the antibody molecule can be determined. In addition, the morphology estimation unit 237 can also evaluate the abnormality of the entire molecule as damage to the structure.

[0092] The morphology estimating unit 237 can use the electron density data obtained by measuring the standard sample with low motility as reference data. Thus, even if there is no known electron density data, the morphology of the constituent elements of the molecule can be estimated using the measurement results of the standard sample.

[0093] The morphology estimation unit 237 may also have an artificial intelligence (AI) function represented by a machine learning model. When the morphology estimation unit 237 has an AI function, the accumulated analytical data can be used as reference data for teacher data. Thus, even in the absence of a standard sample or data that clearly serves as a benchmark, the accumulated analytical data can be used for teacher data, and the morphology of the constituent elements of the molecule to be detected in the object data can be analyzed and evaluated by AI.

[0094] For example, it is possible to utilize a machine learning model composed of a neural network including an input layer, an intermediate layer, and an output layer. When the electron density distribution of the molecule to be tested and its morphological information or quality information are known, they are associated and used as teacher data to enable the machine learning model to learn. That is, the weights of each neuron in the intermediate layer are pre-adjusted by repeatedly learning in a manner that the electron density distribution of the molecule to be tested is input and the correct morphological information or quality information is output with a high probability by the output layer. At this time, in order to be able to roughly identify the object, a convolutional neural network can also be used. Therefore, even if no special parameters are used, as long as the electron density distribution of the object data is input into the machine learning model, the morphology of the molecule to be tested can be inferred or the quality can be determined.

[0095] Preferably, the morphology estimation unit 237 evaluates the difference derived from at least one of conformation and motility specified based on the specified information. Thus, for example, an evaluation result can be obtained for the constituent element of the test molecule for which the difference derived from at least one of conformation and motility is to be evaluated by specifying a range with a mouse.

[0096] The morphology estimation unit 237 may also compare the molecular structure or mobility of the same sample evaluated at a certain time t1 with the molecular structure or mobility at a certain time t2 (t1 < t2) after a certain period of time. Since the difference in at least one of the conformation and mobility caused by the change over time becomes clear, the user can confirm the influence of time.

[0097] The morphology estimation unit 237 can also evaluate the stability of the molecule to be tested. For example, the morphology estimation unit 237 can determine the presence or absence of the constituent elements of the molecule to be tested. In the absence of constituent elements, it is preferred to also estimate the position or size of the gap. For example, the stability of the Fc brought about by the sugar chain modification of the antibody molecule can be evaluated. In addition, the defect of the sugar chain originally bound to the Fc of the antibody molecule can also be determined. The morphology estimation unit 237 can also evaluate the non-equivalence of the constituent elements. For example, the non-equivalence evaluation of the Fab of the antibody molecule can be performed. In this way, in the case where there is a region where the mobility is increased due to the state of the antibody molecule being different from the state of the benchmark, the region can be determined and displayed.

[0098] The quality determination unit 238 determines whether the test molecule meets the quality standard based on the estimated morphology of the test molecule. For example, for a Y-type antibody molecule, it is determined whether the parameters indicated by the estimated morphology are included in the specified value range. The so-called parameters indicated by the morphology may include the size or angle of the constituent elements, or a numerical value indicating the motility. For example, whether the degree of opening of the Fab that estimates the difference in conformation is within the specified value range belongs to this parameter. The specified value is a threshold value based on a standard of reference data or a threshold value containing a confidence limit. For example, if it is a manufacturing process of a Y-type antibody molecule, an antibody that is opened to more than 140°, which is a specified value of the angle formed by the left and right Fabs, is considered to be a deteriorated antibody.

[0099] Furthermore, the quality determination unit 238 can determine the product as a defective product when the parameter is not included in the specified value range. For example, the molecule to be tested is a Y-type antibody molecule, and the constituent element is Fab or Fc. By measuring and analyzing with such a setting, the quality of the product can be determined in the manufacturing process of the Y-type antibody molecule. In addition, instead of using the specified value, information corresponding to the morphology of the molecule to be tested and information on the quality benchmark can be input into the AI-based learning model to determine the quality compliance.

[0100] [Morphological analysis method]

[0101] The operation of the morphological analysis device 200 configured as described above will be described. Figure 4 2 is a flowchart showing a morphological analysis method of motion. First, the morphological analysis device 200 reads reference data and component element identification information (step S1).

[0102] Reference data is, for example, known electron density data of an antibody molecule whose morphology is to be analyzed. The analysis of the morphology includes an evaluation of differences derived from at least one of conformation and mobility. The constituent element determination information is information for determining the portion corresponding to each constituent element from the electron density data of the molecule under test. The information read out may further include information pre-specifying the constituent element whose morphology is to be analyzed.

[0103] Next, the object data is read out (step S2), and the object data is displayed (step S3). The object data is displayed as a projection image in a specified direction. Then, the morphological analysis device 200 accepts input for specifying the constituent elements of the morphology to be analyzed for the molecule under test (step S4). For example, Fab can be cited as a constituent element of an antibody molecule. In the case where no specification is required, it can also be set to be able to skip the specification.

[0104] The morphological analysis device 200 determines whether there is information for specifying the constituent elements (step S5). If there is no specification of the constituent elements, all the constituent elements are determined (step S6). On the other hand, if there is information for specifying the constituent elements, the specified constituent elements are determined (step S7). For example, the Fc, the left Fab, and the right Fab of the antibody molecule are assigned as constituent elements. The details of the determination of the constituent elements will be described later.

[0105] Then, the morphological analysis device 200 uses the state of the electron density of the determined constituent element to infer its morphology (step S8). For example, the electron density distribution of the Fab representing the antibody molecule is sliced ​​with the surface containing the central axis of the Fab, and the longitudinal width and the transverse width on the cross section are detected. The longitudinal direction is the central axis direction, and the transverse direction is the direction perpendicular to the longitudinal direction. And then, the cross-sectional area of ​​the constituent element is roughly calculated according to the longitudinal width and the transverse width of the cross section. The longitudinal width and the transverse width are substituted into the major axis and the minor axis, and the area of ​​the ellipse can be calculated.

[0106] The morphological analysis device 200 may also display a mark indicating the morphology based on the estimated morphology (step S9). A mark indicating the motility may be an arrow mark. A mark indicating the conformational change may be an encircling line mark. The details of each mark will be described later.

[0107] The morphological analysis device 200 determines the estimated result of the morphology (step S10). For example, for the Fab of the antibody molecule, when its cross-sectional area is larger than the reference, it is determined that the motility is high, and when its cross-sectional area is smaller than the reference, it is determined that the motility is low. In addition, it is also possible to confirm that, for example, the sample is not a defective product based on the determination. Then, the determination result is displayed (step S11), and a series of actions are terminated.

[0108] In addition, when evaluating the mobility, the length of the transverse width of the cross section can be compared between the Fab on the left side and the Fab on the right side of the antibody molecule, and the one with the longer transverse width can be evaluated as a Fab with higher mobility, and the one with the shorter transverse width can be evaluated as a Fab with lower mobility. In addition, it can be determined whether there is a difference in mobility between the two Fabs.

[0109] In addition, for example, the direction (ascending degree) of the Fab that forms a Y shape in the Y-shaped antibody molecule can also be assessed. In this case, the angle of the central axis of each Fab relative to the central axis of Fc can be assessed, and the angle formed by each other of Fab can also be assessed. And, when the angle of Fab relative to Fc is below the specified value or the angle formed by each other of Fab is above the specified value, it can be determined that Fab has dropped from the original position and the antibody molecule is close to the T-shaped, and the sample is considered as a defective product. Like this, the opening degree of the Y-shaped antibody molecule can be regarded as an index of the freshness of the assessment sample.

[0110] [Determination of constituent elements]

[0111] Figure 5 (a) and (b) are diagrams showing a projection image of the electron density of target data and identified components, respectively. Figure 5 (a) is a projection image showing the electron density distribution of the object data of the human serum antibody molecule 401. The projection image is an image in which the electron density of the human serum antibody molecule is projected in a certain direction. The human serum antibody molecule 401 represented by the projection image maintains a Y-shape.

[0112] With respect to the distribution of the electron density, the regions of Fab and Fc were detected as components by Y-shaped search. Figure 5 (b) shows the detection results of each area. Figure 5 As shown in (b), the regions of Fab and Fc arranged vertically above and below are detected in the solution, and marks 500a, 500b and 500c (line segments) showing the regions of the left Fab, right Fab and Fc are displayed. Since the projection direction is fixed, the left Fab and the right Fab can be determined.

[0113] [Form of constituent elements]

[0114] (Observation of physical differences arising directly from chemical differences)

[0115] By comparing the reference data with the object data, the form of the constituent elements of the object data can be inferred. The form of the constituent elements can be expressed as a physical difference derived from chemical differences between the molecules determined by the object data and the constituent elements of the molecules determined by the reference data. As the chemical difference, a short-circuit, defect, elongation or addition of certain substances to a part of the structure of the molecule determined by the object data relative to the reference molecule can be cited. In the case of significant chemical changes, the lack of electron density in the short circuit and defect is manifested as a lack of electron density in the part, and the excess electron density is manifested in the elongation and addition of certain substances. For example, in the scenario of developing a new design of bispecific antibodies or the scenario of developing culture conditions, by quickly knowing the addition state of the sugar chain or the modification caused by unexpected chemical substances, it is possible to judge its success or failure, which can improve the efficiency of the process. This physical difference directly derived from chemical differences mainly occurs in the comparison of different objects.

[0116] (Observation of physical differences independent of chemical differences)

[0117] The form of the constituent elements may also be a difference in at least one of the conformation and mobility of the molecule determined by the object data relative to the molecule determined by the reference data. In this case, even if the difference in chemical composition such as a defect is extremely small and no physical difference such as a direct defect in electron density is observed, it may sometimes be caused by at least one of the above-mentioned conformation and mobility as a result due to indirect effects. In this case, there are further cases where the molecule determined by the object data is originally different from the molecule determined by the reference data, and the two are the same thing with only a different comparison time point.

[0118] In addition, examples of the difference derived from at least one of conformation and motility include a difference in conformation, a difference in distribution of conformations, a difference in motility, or a difference in distribution of motility.

[0119] (When the objects of comparison are different)

[0120] When the comparison objects are originally different, it is effective to observe the difference derived from at least one of the conformation and the motility. For example, a molecule determined by the reference data is used as a benchmark and a molecule determined by the object data is observed. Examples include the scenario of designing the antibody drug function itself for the same antibody, or the scenario of analyzing a molecule that produces a different function despite producing the same antibody drug as the benchmark.

[0121] (When the objects of comparison are the same object at different time points)

[0122] Even when the comparison object is the same substance at different time points, it is effective to observe the difference from at least one of the conformation and mobility. For example, the change of the molecular monomer over time can be observed. In this case, the change or distribution of at least one of the conformation and mobility can be observed. It is effective in the formulation development of antibody drugs or the quality management of the manufactured antibody drugs.

[0123] [Display of Marks]

[0124] The form estimated using the state of the electron density of the constituent elements can be indicated by using a symbol. Figure 5 In the example shown in (b), the components of the left Fab, right Fab and Fc are respectively indicated by labels 500a, 500b and 500c (line segments). In this way, the labels can show the conformation of the molecule to be tested. The labels 500a, 500b and 500c can also be indicated by double lines along the molecular chain.

[0125] In addition, the label can also indicate the mobility of the molecule being tested. Figure 6 (a) and (b) are diagrams showing marks (arrows) 501 and 502, respectively. Figure 6 In the example shown in (a), a line segment with arrows at both ends showing the mobility of the left Fab of the human serum antibody molecule 402 is superimposed on the projection image of the electron density of the object data obtained for the human serum antibody molecule 402. The length of the line segment corresponding to the lateral width on the cross section including the central axis of the Fab can represent the magnitude of the mobility, and the arrows at both ends represent the directionality of the movement. According to the mark 501, the left Fab is visualized as having a narrow lateral width and no spatial extension, which clearly shows that its mobility is low.

[0126] exist Figure 6 In the example shown in (b), a line segment with arrows at both ends showing the mobility of the right Fab of the human serum antibody molecule 403 is displayed superimposed on the projection image of the electron density of the target data obtained for the human serum antibody molecule 403. According to the mark 502, the lateral width of the right Fab is significantly larger than that of the left Fab, and the spatial expansion is visualized, which clearly shows that its mobility is high.

[0127] The mark indicating the estimated motility as described above is a three-dimensional vector, and its shape on the screen varies depending on the projection direction of the antibody molecule to be analyzed. Figure 7 (a) to (c) are diagrams showing projection images of electron density and marks (arrows) when the antibody molecule 404 to be analyzed is viewed from directions perpendicular to each other. Figure 7 (b) and (c) show the Figure 7The antibody molecule 404 and the marker 503 are viewed from the directions 7B and 7C shown in FIG. Figure 7 (a) is the right view and top view when the main view is taken as the main view). It can be seen that the Fab region is particularly Figure 7 In the projection image of (c), it appears to be greatly expanded and moves in the direction perpendicular to the surface formed by the Y shape.

[0128] Labels can also be used to indicate physical differences in the constituent elements of a molecule that are directly derived from chemical differences. Examples of physical differences in the constituent elements of a molecule that are directly derived from chemical differences include the cutting or loss of a constituent element. Figure 8 (a) and (b) are diagrams showing marks (encircling lines) 504 and 505, respectively. Figure 8 In the example shown in (a), the sugar chain bound to Fc in antibody molecule 405 is cut, and as a result, the sugar chain in the center of Fc disappears and Fc is cleaved. In the case where the structure of the control data is missing relative to the structure of the reference data, the area can be indicated by an encircling line. Figure 8 In the example shown in (a) of FIG. 1 , a mark 504 formed by a balloon-shaped surrounding line is displayed in superposition with the cracked part of Fc, thereby indicating the position or size of the cracked part.

[0129] In addition, in the example shown, it can be seen that a void is generated in the center of the Fc in the antibody molecule 406. The void indicates that there may be no sugar chains originally bound to the Fc. By displaying a mark 505 formed by a balloon-shaped surrounding line superimposed on the void, the position or size of the void can be shown. In addition, the mark is not limited to an arrow or a surrounding line, and for example, it can also be a geometric display of physical differences directly derived from chemical differences to identify the constituent elements of the molecule, such as a straight line, a curve, and a point (including a starting point and an end point).

[0130] [Application to the manufacturing process of pharmaceuticals]

[0131] The above-described method for analyzing the morphology of a test molecule is particularly suitable for application to the production process of biopharmaceuticals containing proteins as components. Fig. 9 This is a schematic diagram showing the production process of biopharmaceuticals. Biopharmaceuticals are obtained by screening antibody molecules or vectors cultured under predetermined conditions, selecting and purifying the resulting culture strains, and culturing the desired biopharmaceuticals in large quantities.

[0132] In this case, the factors that determine the conditions for the cultivation and purification of raw materials include structural identity (LC, LC / MS) or sugar chain structure as characteristic analysis, its heterogeneity (related to activity) or the completeness of the carrier (stability of the virus capsid), etc., and solutions that do not meet certain quality are excluded. In this way, the idea of ​​"Quality by Design (QbD)" that continuously monitors the manufacturing process to optimize the manufacturing process and design quality becomes important.

[0133] The above-mentioned analysis method of the morphology of the tested molecule is effective in determining whether a certain quality is met. Even users who do not have sufficient insights can quickly and easily perform characteristic analysis, which requires a lot of effort and time for method development in advance in the existing structural analysis method. In addition, it can reduce the cost and human burden of process optimization in exploratory research or quality management.

[0134] In addition, in the above examples, the application of the morphological analysis method to antibody molecules is mainly cited, but it can also be applied to molecules that store / transport substances. Ferritin molecules have a cavity inside, and by opening and closing the hatch part, the substance to be transported can be stored inside. For example, the above-mentioned morphological analysis method can evaluate the opening and closing properties of the hatch part of the capsid molecule and confirm the integrity of the shape.

[0135] [Details of the structure determination unit]

[0136] Fig.10 2 is a block diagram showing the details of the structure identifying unit. The morphology analysis device 200 acquires data of an X-ray scattering distribution map measured by the small-angle X-ray scattering device 100 and identifies the structure of the polymer based on the data.

[0137] The structure identification unit 220 includes a structure model generation unit 221 , a theoretical scattering intensity calculation unit 222 , an index calculation unit 223 , a correlation determination unit 224 , a tendency analysis unit 225 , a theoretical size calculation unit 226 , an actual size calculation unit 227 , a comprehensive determination unit 228 , and a structure model selection unit 229 .

[0138] The structural model generation unit 221 generates a plurality of structural models based on an actually measured X-ray scattering distribution map obtained by measuring the sample S0. The generation of the structural model will be described in detail later.

[0139] Based on the obtained structural model, an index χ indicating the degree of consistency between the measured data for each structural model and the calculated scattering curve can be calculated. 2 and the calculated value Rg(ind) of the dynamic radius of gyration of the electron density model, and a first plot is created by plotting them (see the following Fig.13). The dynamic radius of gyration Rg of a molecule is a preferred example of a parameter representing the size of a molecule, and other parameters representing the size of a molecule may also be used.

[0140] The structure model generation unit 221 preferably generates a plurality of structure models by changing conditions when the distribution of the first drawing has a tendency. Thus, when any of the structure models is not appropriate, it is possible to try to regenerate the structure model. The structure model generation unit 221 can also generate a plurality of structure models according to conditions based on the user's instructions. Thus, when any of the structure models is not appropriate, it is possible to try to generate the structure model by changing conditions.

[0141] The structure model generating unit 221 can generate each structure model in a plurality of structure models one by one by repeated processing according to the setting. Thus, it is possible to suppress useless processing and allocate the required computing resources to promote the processing while confirming the validity of the structure model. In addition, the structure model generating unit 221 can also generate each structure model in a plurality of structure models at one time by parallel processing according to the setting. Thus, it is possible to reproduce the structure model of biopolymers with high validity in a short time. The theoretical scattering intensity calculating unit 222 calculates the theoretical scattering intensity according to each structure model in a plurality of structure models.

[0142] The index calculation unit 223 calculates an index indicating the degree of consistency between the calculated X-ray scattering distribution map calculated based on each of the plurality of structural models and the measured X-ray scattering distribution map. Specifically, as the statistically processed index, x 2 , but any index indicating the degree of consistency with the X-ray scattering distribution diagram is not particularly limited. 2 In addition, the parameters of normal distribution or Poisson distribution, R value, RMS value and RMD value which are indices showing the consistency between the structure factor obtained from the measured diffraction data and the structure factor based on the model structure obtained by analysis, etc. can be cited. 2 It can be calculated as follows.

[0143] [Mathematical formula 3]

[0144]

[0145] The correlation determination unit 224 creates a first plot by plotting a parameter representing the molecular size relative to the above-mentioned calculated index for each of the plurality of structural models, and determines whether the first plot is correlated. Preferably, the correlation determination unit 224 displays the first plot on a display. Thus, the user can visually confirm whether the first plot is correlated.

[0146] When performing iterative processing, the correlation determination unit 224 preferably determines whether there is correlation in the drawings for each iterative processing. Thus, the processing can be advanced while confirming the validity of the structural model for each iterative processing. As a result, the processing can be performed efficiently when computing resources are limited.

[0147] The tendency analysis unit 225 performs multivariate analysis on the distribution of the first drawing when there is no correlation in the first drawing. As a multivariate analysis, for example, PCA (principal component analysis) can be cited. Thus, for example, even if there is no correlation, it is possible to determine whether there is a certain tendency. For example, even if there is no correlation in the overall data, by separating multiple types of correlated data, various data can be used as correlated data.

[0148] The theoretical size calculation unit 226 creates a second plot by plotting parameters representing the molecular size calculated based on a plurality of structural models for each voxel size. As the parameter representing the calculated molecular size, the x-value of the regression line of the calculated value of the gyration radius Rg(ind) is preferably 2 = 1. Then, the theoretical size calculation unit 226 calculates a calculated value Rg(calc) of a parameter indicating the molecular size of the polymer in the solution that is independent of the voxel size, using the second mapping.

[0149] The measured size calculation unit 227 calculates the parameter representing the molecular size of the polymer in the solution as the measured value based on the measured X-ray scattering distribution diagram. Specifically, the measured value Rg (exp) of the dynamic radius of rotation of the molecule can be calculated by performing a Guinier plot. In addition, when the sample forms a hydration circle in the solution, it is preferred to treat the value obtained by subtracting the specified value of the amount of hydration circle from the value obtained by the Guinier plot as the measured value Rg (exp).

[0150] The comprehensive determination unit 228 determines whether there is no statistically significant difference between the calculated value Rg(calc) of the rotation radius calculated by the theoretical size calculation unit 226 and the measured value Rg(exp) calculated by the measured size calculation unit 227. If it is determined that there is no significant difference, a representative structural model is selected from the plurality of structural models. If it is determined that there is a significant difference, the processing is terminated.

[0151] The structural model selection unit 229 selects a representative structural model from a plurality of structural models based on the calculated index. 2The structural model with a value close to 1 is selected as the representative structural model. Since the representative structural model is selected based on the index, the structural model of the polymer in the solution with a dynamically fluctuating structure can be accurately reproduced. As a result, even biopolymers in the solution with no prior information can be accurately visualized.

[0152] The structural model selection unit 229 preferably does not select a representative structural model when there is no correlation with the first drawing. Thus, when any structural model is not appropriate, the determination of the structural model can be stopped, and useless calculations can be omitted. In addition, the structural model can be regenerated or re-tested according to the situation.

[0153] The structural model selection unit 229 preferably does not select a representative structural model if the difference between the calculated value and the measured value is not within a predetermined range. Thus, even if a structural model can be determined, if the structural model cannot guarantee the validity of the molecular size, the determination of the structural model can be stopped.

[0154] [Structure determination method]

[0155] (Overall Method)

[0156] A method of specifying the structure of a polymer in a solution using the morphological analysis system 10 configured as described above will be described. Fig.11 2 is a flow chart showing a structure determination method. First, the small-angle X-ray scattering device 100 delivers a sample S0 in a solution to a predetermined position and irradiates the sample S0 with X-rays. The small-angle X-ray scattering device 100 detects the scattered X-rays and sends them as data of a small-angle X-ray scattering distribution map to the computer 205. The computer 205 stores the received data of the X-ray scattering distribution map in advance.

[0157] The computer 205 receives the designation from the user and reads the data of the X-ray scattering distribution map obtained from the sample S0 whose structure is to be determined (step S1). Then, the computer 205 obtains the generation conditions of the structural model such as the box size, voxel size, or the number of attempts per voxel size designated by the user (step S2).

[0158] According to the obtained generation conditions, a structural model derived from the read X-ray scattering distribution map is generated (step S3). The details of the generation of the structural model will be described later. Next, the theoretical scattering intensity is calculated based on the generated structural model (step S4). Based on the read measured X-ray scattering distribution map and the calculated theoretical scattering intensity, an index x representing the consistency between the calculated X-ray scattering distribution map and the measured X-ray scattering distribution map is calculated. 2 and the calculated value Rg(ind) of the particle's gyration radius for each structural model (step S5), and create χ2 - Plotting of Rg(ind) (step S6).

[0159] Next, as a repetition condition, it is determined whether the generation of the structural model has been attempted a specified number of times for a specific voxel size (step S7). If it is determined that the specified number of attempts have not been made, the process returns to step S3. If it is determined that the specified number of attempts have been made, the process proceeds to step S8. The specified number of attempts is, for example, 50 times. In the above example, in step S7, the process is repeated for a specific voxel size, but the process may be repeated only a specified number of times.

[0160] Determine the x created for a particular voxel size 2 -Whether the drawing of Rg(ind) is correlated (step S8). The details of the correlation determination process will be described later. If it is determined that there is no correlation, it is determined whether the drawing has a certain tendency by multivariate analysis or the like (step S9). If it is determined that there is a tendency, return to step S3, change the conditions and create the structural model again. If it is determined that there is no tendency, the series of processing is terminated without selecting a representative of the structural model. By determining whether there is a correlation and terminating the processing without prospects at the stage where the repeated processing of step S7 is completed, computing resources can be used efficiently.

[0161] On the other hand, if it is determined in step S8 that there is a correlation, it is determined whether the condition for ending the repetition that the structural model is generated for all the multiple voxel sizes is satisfied (step S10). If it is determined that the condition is not satisfied, the voxel size is changed and the process returns to step S3.

[0162] If the condition for ending the repetition is satisfied in step S10, the set of structural models is used to obtain the number of 2 -Rg(ind) is plotted to obtain a regression line, and the χ value of the regression line is calculated. 2 =1 (step S11). Then, a plot of the intercept Rg(sect) with respect to the voxel size is created (step S12), a regression line is obtained from the plot, and a calculated value Rg(calc) of the gyration radius of the molecule is calculated as an extrapolated value in the regression line where the voxel size is zero (step S13). On the other hand, a Guinier plot is created for the read X-ray scattering distribution diagram, and an actual value Rg(exp) of the gyration radius of the molecule is calculated (step S14).

[0163] Next, by determining whether the difference between the calculated value Rg(calc) of the rotation radius and the measured value Rg(exp) is within a certain range, it is determined whether the calculated value Rg(calc) is appropriate (step S15). At this time, when the sample forms a hydration ring in the solution, it is preferred to treat the value obtained by subtracting the specified value of the amount of hydration ring from the value obtained by the Guinier plot as the measured value Rg(exp). When a hydration ring is formed, it is preferred that the specified value is

[0164] If the calculated value Rg(calc) is not determined to be appropriate in step S15, the series of processing is terminated without selecting a representative of the structural model. 2 A representative structural model is selected from the models close to 1, and the selected structural model is output to the output device 290 (step S16), and a series of processing is terminated.

[0165] The structural model represented does not have to be single, but can also be multiple. 2 When the mapping of -Rg(ind) has correlations for each voxel size, there may be representative structural models as many as the number of correlations as a whole. In the above process, only representative structural models as many as the number of correlations may be selected.

[0166] In the above example, each of the plurality of structural models is generated in each repetition, but it is also possible to generate them all at once by parallel processing. In addition, a certain number of parallel processings may be repeated. Which processing is selected can be determined based on which of the computing resources and the speed of obtaining the results is emphasized.

[0167] (Generation of structural model)

[0168] Next, the details of the generation of the structural model are described. When a biopolymer in a solution is irradiated with X-rays, A gentle annular scattered X-ray intensity peak is generated in the range of about . By integrating it in the circumferential direction, an X-ray small-angle scattering distribution diagram can be obtained. In particular, by obtaining the q interval in the range of about 100 to 1000 with high accuracy, By analyzing the scattered intensity data in the range up to 100 Å, the electron density meaningful to the actual structure can be directly visualized.

[0169] Fig.12 is a schematic diagram showing the structural model. In the generation of the structural model, the volume of a cubic box with a side of H that contains the actual space of the polymer is first discretized into an N×N×N grid of cubic voxels (in Fig.12In the example shown, N=4). Fig.12 The density f(x, y, z) of each voxel is randomly given a certain range of values, as shown by the shade of each voxel. Then, the three-dimensional reciprocal lattice space intensity is calculated based on the three-dimensional structure factor, and is divided into concentric shells as a function of the size of the scattering vector q.

[0170] Furthermore, the three-dimensional scattering intensity is converted into a one-dimensional distribution diagram and compared with the experimental scattering data. The three-dimensional structure factor is scaled to be consistent with the experimental data of the concentric shells of each q, and a new electron density map is created in real space by inverse Fourier transformation. The density outside the map is set to zero. The new structure factor is obtained by forward Fourier transformation, and the cycle is repeated until convergence. In this way, a different structural model is generated each time, and the index χ is calculated for each structural model. 2 And the calculated value Rg(ind).

[0171] (Related judgment processing)

[0172] By calculating the χ 2 By plotting Rg(ind) and Rg(ind) to determine whether they are correlated, the validity of the structural model for the measured data can be determined. Whether the plots are correlated can be objectively determined by using, for example, the correlation coefficient. Furthermore, if there is no correlation, by determining whether the plots have a tendency, it can be determined whether the data is hopeless or has the potential to become meaningful depending on the measurement conditions.

[0173] Fig.13 (a) and (b) show the ideal distribution and the unanalyzable χ 2 - Coordinate diagram of the Rg(ind) distribution. Fig.13 In the ideal distribution shown in (a), 2 -Rg(ind) plots show correlation, indicating that multiple structural models are appropriate. Fig.13 In the unanalyzable distribution shown in (b), χ 2 - The plot of Rg(ind) shows no correlation or trend, indicating that many structural models are not appropriate. If the structural model is appropriate, you can choose χ 2 A structural model in a range from 1 to a specified value is included as a representative structural model.

[0174] [Example 1]

[0175] In the process of producing the antibody-drug conjugate, the electron density of the starting antibody molecule and the antibody-drug conjugate was observed using the above-mentioned structural model determination method. Fig.14(a) and (b) are projection images showing the electron density of the raw antibody molecule and the antibody-drug complex, respectively. Both show the projection image of the electron density of the Y-shaped antibody molecule, and it can be confirmed that the raw antibody molecule with a small spread of electron density is changing into an antibody-drug complex with a large spread of electron density. The left Fab, right Fab, and Fc regions of the antibody molecule are shown by line segment marks 600a to 600c.

[0176] In addition, the status of the produced antibody-drug conjugate at day 1 and month 2 was observed using the above-mentioned structural model determination method. Fig.15 (a), (b) are projection images of the electron density of the antibody molecule on the first day after manufacture and the antibody molecule on the second month, respectively. As shown in the arrow marks 601, 602, it can be seen that the mobility of the antibody molecule on the second month and the antibody molecule on the first day has increased. Over time, the Fab originally formed into a Y-shaped shape has drooped, and the antibody molecule is changing from a Y-shaped shape to a very open Y-shaped shape. In addition, it can also be seen that the fluctuation of Fab has also become larger. This is considered to be the result of the aggravated deterioration of the antibody-drug complex after manufacture.

[0177] [Example 2]

[0178] For the Y-type antibody molecule, the electron density data that may become the target data were calculated. 2 The parameter is calculated by weighted average of the observed values ​​weighted by the value of . By evaluating each structural model with a predetermined weight function as the target for all assumed structural models, objective evaluation without arbitrariness can be performed. Fig.16 is Rg(ind)-χ which shows each structural model and its representation 2 Schematic diagram of points on the coordinate graph. The Fab lengths 701a to 705a on the left and 701b to 705b on the right of the Y-shaped antibody molecule respectively represent the lengths of the Fab before correction in each structural model. 2 The structural model closest to 1 is the representative structural model.

[0179] exist Fig.16 In the example shown, according to the χ determined by the structural model 2 The weighting function was evaluated by using the value of . As the weighting function, the mathematical formula (1) was used when a=0, b=4, and n=10 were set. Fig.17 The Fab lengths of the left and right Fabs of the Y-shaped antibody molecules are shown in Figure 1. 2 Table of calculation examples for the corresponding weighting. For the structural models numbered 1 to 50, according to χ 2The weighted corrected observed values ​​of the Fab lengths on the left and right are calculated. In addition, the weighted average lengths and standard deviations σ are calculated by summing the calculated corrected observed values. In this way, the parameters of the molecular shape data of the molecule to be analyzed can be calculated. In addition, in this embodiment, the length of the Fab is used as a parameter, but other parameters such as the opening angle between the arms can also be used.

[0180] Description of Reference Numerals

[0181] 10 Morphological analysis system

[0182] 100 X-ray small angle scattering device

[0183] 110 X-ray generating unit

[0184] 111 X-ray source

[0185] 115 Optical System

[0186] 117 Kratt Block

[0187] 120 Sample loading mechanism

[0188] 125 Sample holding tube

[0189] 130 Detector

[0190] 140 Control Unit

[0191] 200 Morphological Analysis Device

[0192] 205 Computer

[0193] 210 Basic functional units

[0194] 211 Input / output control unit

[0195] 215 Measurement control unit

[0196] 217 Measurement data storage unit

[0197] 220 Structure determination unit

[0198] 221 Structural Model Generation Department

[0199] 222 Theoretical scattering intensity calculation unit

[0200] 223 Index Calculation Department

[0201] 224 Related Judgment Department

[0202] 225 Trend Analysis Department

[0203] 226 Theoretical size calculation unit

[0204] 227 Actual measurement dimension calculation section

[0205] 228 Comprehensive Judgment Department

[0206] 229 Structural Model Selection Department

[0207] 230 Morphological Analysis Unit

[0208] 234 Reference data storage unit

[0209] 235 Object Data Storage Unit

[0210] 236 Output information generation unit

[0211] 237 Morphological Estimation Department

[0212] 238 Quality Assurance Department

[0213] 280 Input Device

[0214] 290 Output device

[0215] 401~406 Human serum antibody molecules

[0216] 500a~500c、600a~600c Mark (line segment)

[0217] 501~503, 601, 602 marks (arrows)

[0218] 504, 505 Mark (enclosing line)

[0219] L Control bus.

Claims

1. A morphological analysis device for analyzing the morphology of molecules in a solution, characterized in that: have: A reference data storage unit storing reference data, wherein the reference data is molecular shape data of a molecule to be detected that is determined in advance; An object data storage unit that stores object data, the object data being molecular shape data of the molecule to be analyzed; as well as a morphology estimating unit for estimating the morphology of the constituent elements of the molecule in the object data by determining the state of the physical quantity of the constituent molecular shape of the object data relative to the reference data, Both the reference data and the object data have a resolution at a molecular level.

2. The morphological analysis device according to claim 1, characterized in that: The morphology of the constituent element is a physical difference directly derived from a chemical difference between the constituent elements of the molecule specified by the object data and the molecule specified by the reference data.

3. The morphological analysis device according to claim 1, characterized in that: The form of the constituent element is a difference in at least one of conformation and mobility between the molecule specified by the object data and the molecule specified by the reference data.

4. The morphological analysis device according to claim 2 or claim 3, characterized in that: The molecule identified by the object data is inherently different from the molecule identified by the reference data.

5. The morphological analysis device according to claim 2 or claim 3, characterized in that: The molecules determined by the subject data are identical to the molecules determined by the reference data at different time points.

6. The morphological analysis device according to any one of claims 1 to 5, characterized in that: The morphology estimating unit determines the state of the physical quantity constituting the molecular shape by the height of the physical quantity constituting the molecular shape or the extended shape or size expressed as a collection thereof.

7. The morphological analysis device according to any one of claims 1 to 5, characterized in that: The morphology estimating unit uses, as the reference data, data of physical quantities constituting the shape of a molecule obtained by measuring a standard sample with low motility.

8. The morphological analysis device according to any one of claims 1 to 5, characterized in that: The morphology estimating unit uses the accumulated analysis data as the reference data for the teacher data.

9. The morphological analysis device according to any one of claims 1 to 5, characterized in that: further comprising an output information generating unit that generates output information in a manner that enables identification of the form of the object data, The output information generating unit displays the object data on an output device, and receives input of designation information for designating a component element of the form to be estimated with respect to the displayed object data. The form estimating unit estimates the form of the component specified based on the specifying information.

10. The morphological analysis device according to any one of claims 1 to 5, characterized in that: further comprising an output information generating unit that generates output information in a manner that enables identification of the form of the object data, The output information generating unit causes an output device to display the object data, and further causes an output device to superimpose a specific mark indicating the estimated form on the object data for display.

11. The morphological analysis device according to any one of claims 1 to 5, characterized in that: The test molecule is a Y-shaped antibody molecule, and the constituent element is Fab or Fc.

12. The morphological analysis device according to claim 11, characterized in that: The device further comprises: an output information generating unit configured to generate output information in a manner that enables identification of the form of the target data; as well as a quality determination unit for determining whether the molecule to be tested meets a quality standard based on the estimated form; The output information generating unit displays that the detected molecule does not meet the quality standard when the detected molecule does not meet the quality standard.

13. A morphological analysis method for analyzing the morphology of molecules in a solution, characterized in that: Include: a step of reading out reference data, wherein the reference data is molecular shape data of the molecule to be detected that is determined in advance; a step of reading out target data, wherein the target data is molecular shape data of the molecule to be analyzed; as well as a step of estimating the form of the constituent elements of the molecule in the object data by determining the state of the physical quantity of the constituent molecular shape of the object data relative to the reference data, Both the reference data and the object data have a resolution at a molecular level.

14. A morphological analysis program for analyzing the morphology of molecules in a solution, characterized in that: Causes the computer to execute: A process of reading out reference data, wherein the reference data is molecular shape data of the molecule to be detected that is determined in advance; a process of reading out target data, the target data being molecular shape data of the molecule to be analyzed; as well as processing of estimating the form of the constituent elements of the molecule in the object data by determining the state of the physical quantity of the constituent molecular shape of the object data relative to the reference data, Both the reference data and the object data have a resolution at a molecular level.

Images