Drug development assistance device, operation method for drug development assistance device, and operation program for drug development assistance device

Abstract

A drug development assistance device comprising a processor, wherein the processor predicts a stable range in which the stability of a large molecule drug is equal to or higher than a set level for a first condition which is one among a plurality of conditions related to the composition of a preservation solution for the large molecule drug.

Description

Drug development support device, drug development support device operation method, and drug development support device operation program

[0001] The technology of the present disclosure relates to a drug development support device, an operating method for a drug development support device, and an operating program for a drug development support device.

[0002] Recently, macromolecular drugs, such as biopharmaceuticals, peptide drugs, and nucleic acid drugs, have attracted attention due to their high efficacy and minimal side effects. For example, biopharmaceuticals use proteins such as interferons and antibodies as their active ingredients. Macromolecular drugs are formulated as injectable solutions dissolved in preservative solutions. Therefore, to maintain the stable quality of macromolecular drugs, it is important to ensure that the preservative solution formulation (also known as formulation) is appropriate for the macromolecular drug. The preservative solution is composed of a base buffer solution and additives such as salts, sugars, amino acids, and surfactants. The preservative solution formulation is determined by several factors, such as the pH, the type and amount of buffer, and the type and amount of additives.

[0003] Conventionally, when determining the formulation of a preservation solution, multiple types of preservation solutions are actually prepared by varying the above-mentioned multiple conditions, and the storage stability of each preservation solution is confirmed through testing. In order to efficiently conduct such testing, a method using Bayesian optimization has been proposed, as described in [Kouta Matsui et al., "Fundamentals of Bayesian Optimization and Its Application to Materials Engineering," Material, Vol. 58, No. 1, pp. 12-16, 2019] (hereinafter referred to as Non-Patent Document 1).

[0004] Conventional methods, including the Bayesian optimization method described in Non-Patent Document 1, do not take into account the inherent stability of macromolecular pharmaceuticals. As a result, when determining the formulation of a preservation solution for a macromolecular pharmaceutical that has a relatively high inherent stability and can withstand slight changes in conditions, many unnecessary tests are often performed. Conversely, when determining the formulation of a preservation solution for a macromolecular pharmaceutical that has a relatively low inherent stability and is sensitive to even slight changes in conditions, sufficient tests are often not performed.

[0005] One embodiment of the technology disclosed herein provides a pharmaceutical development support device capable of conducting tests adapted to the inherent stability of macromolecular pharmaceuticals, an operating method for the pharmaceutical development support device, and an operating program for the pharmaceutical development support device.

[0006] The pharmaceutical development support device disclosed herein includes a processor, which predicts a stability range in which the stability of the polymeric pharmaceutical is at or above a set level for a first condition, which is one of multiple conditions related to the formulation of a preservative solution of the polymeric pharmaceutical.

[0007] The processor preferably derives a search range for testing conditions other than the first condition based on the stability range, the search range depending on the width of the stability range.

[0008] When the stability range is relatively wide, the search range is preferably narrower than when the stability range is relatively narrow.

[0009] The search range is preferably narrowed by reducing at least one of the variations in the types and amounts of the components of the preservation solution.

[0010] The processor preferably presents the search range to the user.

[0011] It is preferable that the processor acquires sequence information of the amino acids that make up the amino acid-derived substance contained in the polymeric pharmaceutical, converts the sequence information into a first feature that represents the characteristics of the substance, integrates the first feature and the first condition to form a second feature, derives a stability index value from the second feature that represents the stability of the polymeric pharmaceutical against the first condition, and derives a stability range based on the stability index value.

[0012] It is preferable that the processor obtains a plurality of stability index values ​​corresponding to a plurality of first conditions by repeatedly integrating the first characteristic quantity and the first condition to obtain a second characteristic quantity while changing the first condition, and deriving a stability index value from the second characteristic quantity, and derives a stability range based on an approximation curve of the plurality of stability index values.

[0013] The processor preferably converts the sequence information into the first feature using a language model that treats the sequence information as a sentence in natural language processing.

[0014] It is preferable that the processor extracts a descriptor representing a physical property of the substance from the sequence information and treats the descriptor as the first feature.

[0015] Preferably, the processor generates the second feature by adding or multiplying all elements of the first feature by the first condition.

[0016] Preferably, the processor generates the second feature by adding the first condition as a new element to the first feature.

[0017] Preferably, the processor inputs the second feature amount to a machine learning model and causes the machine learning model to output a stability index value.

[0018] Preferably, the substance is one of a protein, a peptide, and a nucleic acid.

[0019] Preferably, the protein is an antibody.

[0020] The first condition is preferably the hydrogen ion exponent.

[0021] The operating method of the pharmaceutical development support device disclosed herein includes predicting a stability range in which the stability of the polymeric pharmaceutical is at or above a set level for a first condition, which is one of multiple conditions related to the formulation of a preservative solution of the polymeric pharmaceutical.

[0022] The operating program of the pharmaceutical development support device disclosed herein causes a computer to execute processing including predicting a stability range in which the stability of a polymeric pharmaceutical is at or above a set level for a first condition, which is one of multiple conditions related to the prescription of a preservative solution of the polymeric pharmaceutical.

[0023] According to the technology disclosed herein, it is possible to provide a pharmaceutical development support device capable of conducting tests adapted to the inherent stability of high molecular weight pharmaceuticals, an operating method for the pharmaceutical development support device, and an operating program for the pharmaceutical development support device.

[0024] 1 is a diagram showing a drug development support system. FIG. 1 is a diagram showing the formation of a biopharmaceutical. FIG. 2 is a diagram showing sequence information. FIG. 3 is a diagram showing a search range. FIG. 4 is a block diagram showing a computer constituting a drug development support device and a user terminal. FIG. 5 is a block diagram showing a processing unit of a CPU of the drug development support device. FIG. 6 is a diagram showing a first condition. FIG. 7 is a block diagram showing a detailed configuration of a stability range prediction unit. FIG. 8 is a diagram showing the formation of a conversion model. FIG. 9 is a diagram showing the processing of a conversion unit. FIG. 10 is a diagram showing the processing of an integration unit. FIG. 11 is a diagram showing an integrated vector data group. FIG. 12 is a diagram showing the processing of a first derivation unit. FIG. 13 is a diagram showing a group of stability index values. FIG. 14 is a diagram showing the processing in the learning phase of a derivation model. FIG. 15 is a diagram showing the processing of a second derivation unit. FIG. 16 is a block diagram showing a sequence information input screen. FIG. 17 is a diagram showing a search range display screen. FIG. 18 is a flowchart showing the processing procedure of the drug development support device. FIG. 19 is a diagram showing another example of the processing of the integration unit. FIG. 20 is a diagram showing yet another example of the processing of the integration unit. FIG. 21 is a diagram showing the processing of an extraction unit of a second embodiment. FIG. 22 is a diagram showing a peptide drug or a nucleic acid drug using a peptide or a nucleic acid as an amino acid-derived substance.

[0025] [First embodiment] As an example, as shown in Figure 1, a drug development support system 10 is a system that supports the development of a biopharmaceutical 11, and includes a drug development support device 12 and a user terminal 13. The drug development support device 12 and the user terminal 13 are connected via a network 14. The network 14 is, for example, a WAN (Wide Area Network) such as the Internet or a public communication network.

[0026] The user terminal 13 is installed at a pharmaceutical company developing the biopharmaceutical 11 or at an organization contracted by a pharmaceutical company to develop the biopharmaceutical 11, i.e., a contract research organization (CRO). The user terminal 13 is operated by a user U involved in the development of the biopharmaceutical 11 at the pharmaceutical company or contract research organization (hereinafter collectively referred to as a pharmaceutical facility). The biopharmaceutical 11 is an example of a "polymer drug" according to the technology disclosed herein. Here, a polymer drug is a drug whose active ingredient is an amino acid-derived substance with a molecular weight of 500 or more. Note that while only one user terminal 13 is connected to the pharmaceutical development support device 12 in FIG. 1 , in reality, multiple user terminals 13 at multiple pharmaceutical facilities are connected to the pharmaceutical development support device 12.

[0027] As shown in FIG. 2 as an example, the biopharmaceutical 11 is a mixed solution of an antibody 16, which is an active ingredient, and a preservation solution 17. The antibody 16 is an example of an "amino acid-derived substance" and a "protein" according to the technology of the present disclosure. The preservation solution 17 is composed of a base buffer solution 18 and an additive 19 added to the buffer solution 18. The buffer solution 18 and the additive 19 are examples of "preservation solution components" according to the technology of the present disclosure. The additive 19 is a salt, a sugar, an amino acid, and a surfactant. The preservation solution 17 is prepared with reference to prescription information 20, which specifies the hydrogen ion exponent (pH: Potential Hydrogen), the type and amount of the buffer solution 18, and the types and amounts of the salt, sugar, amino acid, and surfactant. The prescription information 20 may also include temperature.

[0028] There are several types of buffer solutions 18, including phosphate buffer solutions, acetate buffer solutions, and citrate buffer solutions. Examples of salts include sodium chloride, potassium chloride, sodium acetate, and ammonium sulfate. Examples of sugars include pure sugars such as refined white sugar, sucrose, and trehalose, and sugar alcohols such as glycerin and sorbitol. Examples of amino acids include proline, arginine, glutamine, and histidine. Examples of surfactants include polysorbate 20 and polysorbate 80.

[0029] Returning to FIG. 1 , the user terminal 13 transmits a distribution request 15 to the drug development support device 12. The distribution request 15 is a request to have the drug development support device 12 distribute information useful for promoting the development of the biopharmaceutical 11. The distribution request 15 includes sequence information 21. The sequence information 21 is information representing the sequence of amino acids constituting the antibody 16 contained in the biopharmaceutical 11. The sequence information 21 is identified through experiments. Here, the biopharmaceutical 11 is currently under development at a pharmaceutical facility, and the prescription information 20 for an appropriate preservative solution 17 is unknown. Although not shown, the distribution request 15 also includes a terminal ID (identification data) and the like for uniquely identifying the user terminal 13 that sent the distribution request 15.

[0030] When receiving the distribution request 15, the pharmaceutical development support device 12 derives a search range 22 for testing conditions related to the formulation of a preservative solution 17 suitable for the biopharmaceutical 11, as information useful for promoting the development of the biopharmaceutical 11. The search range 22 is then distributed to the user terminal 13 that sent the distribution request 15. When the search range 22 is received, the user terminal 13 makes the search range 22 available for viewing by the user U. Note that in this specification, "test" includes not only an actual test, but also a virtual test such as a prediction using artificial intelligence (AI) or a molecular dynamics (MD) simulation.

[0031] As an example, as shown in FIG. 3 , the sequence information 21 includes a drug ID for uniquely identifying the biopharmaceutical 11. The sequence information 21 describes the order of peptide bonds of the amino acids that make up the antibody 16 contained in the biopharmaceutical 11, from the amino terminus to the carboxyl terminus, using single-letter alphabetic abbreviations representing the amino acids. Since there are approximately 450 amino acids that make up the antibody 16, the sequence information 21 also contains a string of approximately 450 letters. For example, "E" represents glutamic acid, "L" represents leucine, and "G" represents glycine. Such an amino acid sequence is also called a primary structure.

[0032] As an example, as shown in FIG. 4 , the search range 22 includes a drug ID, just like the sequence information 21. The search range 22 is a registered range for storage stability testing of the type and amount of buffer solution 18, as well as the types and amounts of salt, sugar, amino acids, and surfactants. FIG. 4 illustrates a search range 22 in which type 1 and amounts 1 to 3 are registered for each of buffer solution 18, salt, sugar, amino acids, and surfactants. The larger the number, the larger the amount. The type and amount of buffer solution 18, as well as the types and amounts of salt, sugar, amino acids, and surfactants, are examples of "other conditions" according to the technology of the present disclosure.

[0033] 5, the computers that make up the pharmaceutical development support device 12 and the user terminal 13 basically have the same configuration, and include a storage 25, a memory 26, a CPU (Central Processing Unit) 27, a communication unit 28, a display 29, and an input device 30. These are interconnected via a bus line 31.

[0034] The storage 25 is a hard disk drive built into the computer that constitutes the pharmaceutical development support device 12 and the user terminal 13, or connected via a cable or network. Alternatively, the storage 25 is a disk array consisting of multiple hard disk drives. The storage 25 stores control programs such as an operating system, various application programs (hereinafter referred to as APs (Application Programs)), and various data associated with these programs. Note that a solid state drive may be used instead of a hard disk drive.

[0035] The memory 26 is a work memory for the CPU 27 to execute processing. The CPU 27 loads programs stored in the storage 25 into the memory 26 and executes processing in accordance with the programs. In this way, the CPU 27 comprehensively controls each part of the computer. The CPU 27 is an example of a "processor" according to the technology of the present disclosure. The memory 26 may be built into the CPU 27.

[0036] The communication unit 28 is a network interface that controls the transmission of various information via the network 14, etc. The display 29 displays various screens. The various screens are equipped with operation functions using a GUI (Graphical User Interface). The computers that make up the pharmaceutical development support device 12 and the user terminal 13 accept input of operation instructions from an input device 30 via the various screens. The input device 30 is a keyboard, a mouse, a touch panel, a microphone for voice input, etc.

[0037] In the following explanation, the parts of the computer that make up the pharmaceutical development support device 12 (storage 25 and CPU 27) are distinguished by adding the suffix "A" to their reference symbols, and the parts of the computer that make up the user terminal 13 (storage 25, CPU 27, display 29, and input device 30) are distinguished by adding the suffix "B" to their reference symbols.

[0038] 6, an operating program 35 is stored in storage 25A of drug development support device 12. Operating program 35 is an AP for causing a computer to function as drug development support device 12. In other words, operating program 35 is an example of an "operating program for a drug development support device" according to the technology of the present disclosure. Storage 25A also stores a conversion model 36, a first condition 37, a derived model 38, range correspondence information 39, and the like.

[0039] When the operating program 35 is started, the CPU 27A of the computer constituting the pharmaceutical development support device 12 works in cooperation with the memory 26 and the like to function as a request receiving unit 45, a read / write (hereinafter abbreviated as RW (Read Write)) control unit 46, a stability range prediction unit 47, a search range derivation unit 48, and a screen distribution control unit 49.

[0040] The request receiving unit 45 receives various requests from the user terminal 13. In particular, the request receiving unit 45 receives a delivery request 15 from the user terminal 13. The delivery request 15 includes the array information 21 as described above. Therefore, by receiving the delivery request 15, the request receiving unit 45 acquires the array information 21. When the delivery request 15 is received, the request receiving unit 45 outputs the array information 21 included in the delivery request 15 to the RW control unit 46. Furthermore, the request receiving unit 45 outputs the terminal ID of the user terminal 13 included in the delivery request 15 to the screen delivery control unit 49.

[0041] The RW control unit 46 controls the storage of various data in the storage 25A and the reading of various data from the storage 25A. For example, the RW control unit 46 stores the array information 21 from the request receiving unit 45 in the storage 25A. The RW control unit 46 also reads the array information 21 from the storage 25A and outputs the read array information 21 to the stability range prediction unit 47.

[0042] The RW control unit 46 reads the transformation model 36, the first condition 37, and the derived model 38 from the storage 25A, and outputs the read transformation model 36, the first condition 37, and the derived model 38 to the stable range prediction unit 47. The RW control unit 46 also reads the range correspondence information 39 from the storage 25A, and outputs the read range correspondence information 39 to the search range derivation unit 48.

[0043] The stability range prediction unit 47 predicts a stability range 55 in which the stability of the biopharmaceutical 11 is at or above a set level for the first condition 37. The first condition 37 is one of multiple conditions related to the formulation of the preservative solution 17, along with the type and amount of the buffer solution 18 and the types and amounts of salt, sugar, amino acid, and surfactant. Specifically, as shown in FIG. 7 , the first condition 37 is a seven-stage hydrogen ion exponent that varies in increments of 0.5 between 5.0 and 8.0. The stability range prediction unit 47 outputs the predicted stability range 55 to the search range derivation unit 48.

[0044] The search range derivation unit 48 derives the search range 22 according to the width of the stable range 55 based on the range correspondence information 39. The search range derivation unit 48 outputs the derived search range 22 to the screen distribution control unit 49.

[0045] The screen distribution control unit 49 controls the distribution of various screens to the user terminal 13. Specifically, the screen distribution control unit 49 distributes and outputs various screens to the user terminal 13 that has sent the various requests in the form of screen data for web distribution created using a markup language such as XML (Extensible Markup Language). At this time, the screen distribution control unit 49 identifies the user terminal 13 that has sent the various requests based on the terminal ID from the request receiving unit 45. The various screens include an array information input screen 100 (see FIG. 20) for inputting array information 21 and a search range display screen 105 (see FIG. 21) for displaying the search range 22. Note that other data description languages, such as JSON (Javascript (registered trademark) Object Notation), may be used instead of XML.

[0046] As an example, as shown in FIG. 8 , the stable range prediction unit 47 includes a conversion unit 60, an integration unit 61, a first derivation unit 62, and a second derivation unit 63. The conversion unit 60 receives the array information 21 and the conversion model 36 from the RW control unit 46. The conversion unit 60 converts the array information 21 into vector data 65 using the conversion model 36. The vector data 65 is an example of a "first feature" according to the technology of the present disclosure. The conversion unit 60 outputs the vector data 65 to the integration unit 61.

[0047] The integration unit 61 receives the first condition 37 from the RW control unit 46 and the vector data 65 from the conversion unit 60. The integration unit 61 integrates the vector data 65 and the first condition 37 to generate integrated vector data 66 (see FIG. 11 ). More specifically, the integration unit 61 integrates the vector data 65 with each of the seven levels of hydrogen ion exponent of the first condition 37 to generate seven pieces of integrated vector data 66. The integrated vector data 66 is an example of a "second feature amount" according to the technology of the present disclosure. The integration unit 61 outputs an integrated vector data group 66G, which is a collection of the seven pieces of integrated vector data 66, to the first derivation unit 62.

[0048] The first derivation unit 62 receives the derivation model 38 from the RW control unit 46 and the integrated vector data group 66G from the integration unit 61. The first derivation unit 62 uses the derivation model 38 to derive a stability index value 67 (see FIG. 13 ) representing the stability of the biopharmaceutical 11 under the first condition 37 from the integrated vector data 66. More specifically, the first derivation unit 62 derives a stability index value 67 from each of the seven integrated vector data 66. Thus, seven stability index values ​​67 are derived. The first derivation unit 62 outputs a stability index value group 67G, which is a collection of the seven stability index values ​​67, to the second derivation unit 63.

[0049] The second derivation unit 63 receives the stability index value group 67G from the first derivation unit 62. The second derivation unit 63 derives the stability range 55 based on the seven stability index values ​​67 in the stability index value group 67G. The second derivation unit 63 outputs the stability range 55 to the search range derivation unit 48.

[0050] 9, a vectorization unit 71, which is part of a language model 70, is diverted to the conversion model 36. Therefore, it can be said that the conversion unit 60 converts the array information 21 into vector data 65 using the language model 70.

[0051] The language model 70 is a machine learning model originally used in the field of natural language processing (NLP). The language model 70 handles sequence information 21 of amino acids constituting the antibody 16 as a sentence in natural language processing. The language model 70 is based on, for example, BERT (Bidirectional Encoder Representations from Transformers) using a transformer encoder.

[0052] The vectorization unit 71 converts the sequence information 21 input to the language model 70 into vector data 65. The vector data 65 reflects not only the sequence (primary structure) of the amino acids that make up the antibody 16, but also the three-dimensional structure (secondary structure, tertiary structure, and quaternary structure). The vectorization unit 71 outputs the vector data 65 to the processing unit 72. The processing unit 72 outputs a processing result 73 based on the vector data 65.

[0053] The language model 70 is generated by pre-training the pre-trained language model 70A and then fine-tuning the pre-trained language model 70B. The pre-training includes MLM (Masked Language Modeling) and NSP (Next Sentence Prediction). MLM is a learning method that masks a portion of the amino acid sequence in the sequence information 21 and predicts which amino acid will be inserted in the masked portion, a so-called fill-in-the-blank problem. NSP is a learning method that determines whether there is a correlation between the amino acid sequence information 21 that constitutes two different antibodies 16. Fine-tuning is learning according to a desired processing task to be performed by the processing unit 72. For example, the processing task may involve outputting an appropriate temperature value for the preservation solution 17 as a processing result 73 in response to input of the sequence information 21. The vectorization unit 71 of the language model 70 generated in this manner is stored in the storage 25A as the conversion model 36. The pre-learning and fine-tuning may be performed in the drug development support device 12, or may be performed in a device separate from the drug development support device 12. Furthermore, the pre-learning and fine-tuning may be continued even after the conversion model 36 is stored in the storage 25A. Alternatively, only the pre-learning may be performed without fine-tuning. Furthermore, NSP may not be performed.

[0054] As an example, as shown in Figure 10, the conversion unit 60 inputs the sequence information 21 to the conversion model 36 and outputs vector data 65 from the conversion model 36. The vector data 65 includes a drug ID, just like the sequence information 21. The vector data 65 is multidimensional data consisting of a series of real values ​​between 0 and 1, for example. Hereinafter, the real values ​​that make up the vector data 65 will be referred to as elements, and the number of elements will be referred to as the number of dimensions.

[0055] 11 , the integration unit 61 generates integrated vector data 66 by adding the first condition 37 to all elements of vector data 65. The numbers 1 to N in the vector data 65 and the integrated vector data 66 are the numbers of each element of the vector data 65 and the integrated vector data 66. N is a natural number of 2 or greater, and is nothing other than the number of elements of the vector data 65 and the integrated vector data 66, i.e., the number of dimensions. N is, for example, 512, 1024, or 2048.

[0056] 11 illustrates an example in which 5.0, one of the seven hydrogen ion exponents, is added to all elements of vector data 65 to generate integrated vector data 66. The integration unit 61 similarly generates integrated vector data 66 for the remaining six of the seven hydrogen ion exponents. In this way, an integrated vector data group 66G, which is a collection of seven pieces of integrated vector data 66, can be obtained, as shown in FIG. 12 as an example.

[0057] As an example, as shown in FIG. 13 , the first derivation unit 62 inputs integrated vector data 66 to a derivation model 38 and causes the derivation model 38 to output a stability index value 67. The derivation model 38 is configured by a lasso regression model, a LightGBM (Gradient Boosting Machine), a neural network, or the like. The derivation model 38 is an example of a "machine learning model" according to the technology of the present disclosure. The stability index value 67 includes the hydrogen ion exponent of the first condition 37 integrated into the integrated vector data 66. The stability index value 67 ranges from 0, indicating the lowest stability, to 100, indicating the highest stability.

[0058] 13 illustrates an example of deriving a stability index value 67 for 5.0, which is one of the seven hydrogen ion exponents. The first derivation unit 62 similarly derives stability index values ​​67 for the remaining six of the seven hydrogen ion exponents. In this way, a stability index value group 67G, which is a set of seven stability index values ​​67, can be obtained, as shown in FIG. 14 as an example.

[0059] As an example, as shown in FIG. 15 , the derived model 38 is trained by being given learning data (also referred to as teacher data or training data) 80. The learning data 80 is a set of integrated learning vector data 66L and a correct answer index value 67CA. The integrated learning vector data 66L is the integrated vector data 66 of a previously developed biopharmaceutical 11. The correct answer index value 67CA is the stability index value of the previously developed biopharmaceutical 11 relative to the hydrogen ion exponent of the first condition 37 integrated into the integrated learning vector data 66L. The correct answer index value 67CA is, so to speak, data for checking the answer.

[0060] In the learning phase, the integrated vector data for learning 66L is input to the derivation model 38, which in turn outputs a stability index value for learning 67L. This stability index value for learning 67L is compared with the accuracy index value 67CA, and a loss calculation is performed for the derivation model 38 using a loss function according to the comparison result. Then, various coefficients of the derivation model 38 are updated according to the result of the loss calculation, and the derivation model 38 is updated according to the update setting.

[0061] During the learning phase, the series of processes described above, including input of the training integrated vector data 66L to the derived model 38, output of the training stability index value 67L from the derived model 38, loss calculation, update setting, and update of the derived model 38, are repeatedly performed while the training data 80 is exchanged. The repetition of the series of processes described above is terminated when the derivation accuracy of the training stability index value 67L reaches a predetermined set level. The derived model 38 whose derivation accuracy has reached the set level is stored in the storage 25A and used in the first derivation unit 62. Note that learning may be terminated when the series of processes described above has been repeated a set number of times, regardless of the derivation accuracy of the training stability index value 67L. Training of the derived model 38 may be performed in the pharmaceutical development support device 12 or in a device separate from the pharmaceutical development support device 12. Training of the derived model 38 may also be continued even after the derived model 38 is stored in the storage 25A.

[0062] As shown in the graph 85 shown in FIG. 16 , which has the pH on the horizontal axis and the stability index value 67 on the vertical axis, the search range derivation unit 48 derives an approximate curve 87 connecting plots 86 of a plurality of stability index values ​​67 using a well-known technique such as the least squares method. The search range derivation unit 48 calculates the pH at two intersections 89 between the approximate curve 87 and a line 88 passing through 80% of the maximum value of the approximate curve 87. The difference between the calculated pH values ​​is then calculated as the stability range 55. The 80% value of the maximum value of the approximate curve 87 is an example of the "set level" according to the technology of the present disclosure. FIG. 16 illustrates a case where the pH values ​​at the two intersections 89 are 5.68 and 7.73, and 7.73 - 5.68 = 2.05 is calculated as the stability range 55.

[0063] As an example, as shown in FIG. 17 , the range correspondence information 39 is information in which a search range 22 corresponding to a stability range 55 is registered. When the stability range 55 is greater than or equal to 0 and less than 0.5, the search range 22 is types 1 to 3 and amounts 1 to 5 for the buffer solution 18, salt, sugar, amino acid, and surfactant. When this stability range 55 is greater than or equal to 0 and less than 0.5, the search range 22 is all possible variations of type and amount. In other words, when the stability range 55 is greater than or equal to 0 and less than 0.5, the search range 22 is the widest.

[0064] When the stability range 55 is equal to or greater than 0.5 and less than 1.0, the search range 22 includes buffer solution 18, salt, sugar, amino acid, and surfactant in types 1 and 2 and amounts 1 to 5. In other words, when the stability range 55 is equal to or greater than 0.5 and less than 1.0, the variety of types is reduced compared to when the stability range 55 is equal to or greater than 0 and less than 0.5.

[0065] When the stability range 55 is equal to or greater than 2.0 and less than 2.5, the search range 22 contains 1 type and 1 to 3 amounts of buffer solution 18, salt, sugar, amino acid, and surfactant. That is, the search range 22 when the stability range 55 is equal to or greater than 2.0 and less than 2.5 has fewer variations in type and amount compared to the search range 22 when the stability range 55 is equal to or greater than 0 and less than 0.5 and the search range 22 when the stability range 55 is equal to or greater than 0.5 and less than 1.0. In this way, the search range 22 when the stability range 55 is relatively wide is narrower than the search range 22 when the stability range 55 is relatively narrow. Furthermore, the search range 22 is narrowed by reducing at least one of the variations in type and amount of buffer solution 18, salt, sugar, amino acid, and surfactant.

[0066] 18, the second derivation unit 63 derives the search range 22 corresponding to the stable range 55 from the first derivation unit 62 from the range correspondence information 39. In FIG. 18, the case where the stable range 55 is 2.05 is illustrated, as in the example shown in FIG. 16.

[0067] As an example, as shown in Figure 19, a support AP 95 is stored in the storage 25B of the user terminal 13. The support AP 95 is installed in the user terminal 13 by the user U. The support AP 95 is an AP for receiving the development support service for the biopharmaceutical 11 provided by the pharmaceutical development support device 12. When the support AP 95 is activated, the CPU 27B of the user terminal 13 functions as a browser control unit 97 in cooperation with the memory 26 and the like. The browser control unit 97 controls the operation of the dedicated web browser for the support AP 95.

[0068] The browser control unit 97 reproduces various screens based on various screen data from the drug development support device 12 and displays the reproduced various screens on the display 29B. The browser control unit 97 also accepts various operation instructions input by the user U from the input device 30B via the various screens. The browser control unit 97 transmits various requests, including the distribution request 15, to the drug development support device 12 in response to the operation instructions.

[0069] When the support AP 95 is started, a sequence information input screen 100, as shown in Fig. 20 as an example, is displayed on the display 29B under the control of the browser control unit 97. The sequence information input screen 100 is provided with an input box 101 for the sequence information 21. The input box 101 allows the sequence information 21 to be written or a file of the sequence information 21 to be dropped therein.

[0070] After inputting the desired sequence information 21 into the input box 101, the user U selects the start button 102. When the start button 102 is selected, the browser control unit 97 generates a distribution request 15 including the sequence information 21 input into the input box 101 and transmits the generated distribution request 15 to the drug development support device 12.

[0071] Furthermore, when the search range 22 is derived in the pharmaceutical development support device 12, a search range display screen 105 shown in FIG. 21 is displayed on the display 29B under the control of the browser control unit 97. The search range 22 is displayed on the search range display screen 105. In this manner, the search range 22 is presented to the user U in the form of screen data of the search range display screen 105 being distributed by the screen distribution control unit 49.

[0072] A sequence information display button 106 is provided at the top of the search range display screen 105. When the sequence information display button 106 is selected, a display screen for the sequence information 21 is popped up. Furthermore, a save button 107 and an OK button 108 are provided at the bottom of the search range display screen 105. When the save button 107 is selected, the search range 22 is associated with the drug ID and stored in the storage 25B. When the OK button 108 is selected, the display of the search range display screen 105 is cleared.

[0073] Next, the operation of the above configuration will be described with reference to the flowchart shown in Fig. 22 as an example. When the operating program 35 is started in the drug development support device 12, the CPU 27A functions as a request receiving unit 45, a RW control unit 46, a stability range prediction unit 47, a search range derivation unit 48, and a screen distribution control unit 49, as shown in Fig. 6. When the support AP 95 is started in the user terminal 13, the CPU 27B functions as a browser control unit 97, as shown in Fig. 19.

[0074] 20 is displayed on the display 29B of the user terminal 13 under the control of the browser control unit 97. When the user U inputs desired sequence information 21 into the input box 101 on the sequence information input screen 100 and selects the start button 102, a distribution request 15 is transmitted from the browser control unit 97 to the drug development support device 12. As shown in FIG. 1, the distribution request 15 includes the sequence information 21, the terminal ID of the user terminal 13, and the like.

[0075] In the pharmaceutical development support device 12, the request receiving unit 45 receives the distribution request 15, thereby acquiring the sequence information 21 included in the distribution request 15 (YES in step ST100). The sequence information 21 included in the distribution request 15 is output from the request receiving unit 45 to the RW control unit 46 and stored in the storage 25A under the control of the RW control unit 46 (step ST110). In addition, the terminal ID of the user terminal 13 included in the distribution request 15 is output from the request receiving unit 45 to the screen distribution control unit 49.

[0076] The array information 21 is read from the storage 25A by the RW control unit 46 (step ST120). The array information 21 is output from the RW control unit 46 to the stable range prediction unit 47. The RW control unit 46 also reads the transformation model 36, the first condition 37, and the derived model 38 from the storage 25A, and outputs the read transformation model 36, the first condition 37, and the derived model 38 to the stable range prediction unit 47. The RW control unit 46 also reads the range correspondence information 39 from the storage 25A, and outputs the read range correspondence information 39 to the search range derivation unit 48.

[0077] 10, in the conversion unit 60 of the stable range prediction unit 47, the array information 21 is input to the conversion model 36. As a result, vector data 65 is output from the conversion model 36. In this way, the array information 21 is converted into the vector data 65 using the conversion model 36 (step ST130). The vector data 65 is output from the conversion unit 60 to the integration unit 61.

[0078] 11, the integration unit 61 integrates the vector data 65 and the hydrogen ion exponent of the first condition 37 to generate integrated vector data 66 (step ST140). The process of step ST140 is repeated until it has been performed for all levels of hydrogen ion exponent of the first condition 37 (NO in step ST150). When the process of step ST140 has been performed for all levels of hydrogen ion exponent (YES in step ST150), an integrated vector data group 66G, which is a collection of multiple integrated vector data 66 shown in FIG. 12, is output from the integration unit 61 to the first derivation unit 62.

[0079] In the first derivation unit 62, as shown in Fig. 13, the integrated vector data 66 is input to the derivation model 38. As a result, a stability index value 67 is output from the derivation model 38 (step ST160). The process of step ST160 is repeated until it has been performed on all of the integrated vector data 66 (NO in step ST170). When the process of step ST160 has been performed on all of the integrated vector data 66 (YES in step ST170), a stability index value group 67G, which is a collection of multiple stability index values ​​67 shown in Fig. 14, is output from the first derivation unit 62 to the second derivation unit 63.

[0080] 16, the second derivation unit 63 derives the stable range 55 based on the approximation curve 87 of the plurality of stability index values ​​67 (step ST180). The stable range 55 is output from the second derivation unit 63 to the search range derivation unit 48.

[0081] 18, the search range derivation unit 48 derives the search range 22 corresponding to the stable range 55 from the range correspondence information 39 (step ST190). The search range 22 is output from the search range derivation unit 48 to the screen distribution control unit 49.

[0082] The screen distribution control unit 49 generates screen data for the search range display screen 105 shown in Fig. 21 based on the search range 22. Under the control of the screen distribution control unit 49, the screen data for the search range display screen 105 is distributed to the user terminal 13 that is the sender of the distribution request 15 (step ST200).

[0083] In the user terminal 13, under the control of the browser control unit 97, the screen data of the search range display screen 105 is reproduced, and the reproduced search range display screen 105 is displayed on the display 29B. In this way, the search range 22 is presented to the user U. The user U performs a test using the presented search range 22. Note that it is up to the discretion of the user U whether to adopt the presented search range 22 as is or to make some changes based on his or her own knowledge.

[0084] As described above, the stability range prediction unit 47 of the CPU 27A of the pharmaceutical development support device 12 predicts the stability range 55 within which the stability of the biological drug 11 is equal to or exceeds a set level for the first condition 37, which is one of multiple conditions related to the formulation of the preservative solution 17 for the biological drug 11. Predicting the stability range 55 allows the inherent stability of the biological drug 11 to be estimated. As a result, it becomes possible to perform tests adapted to the inherent stability of the biological drug 11. Therefore, when determining the formulation of the preservative solution 17 for the biological drug 11, which has a relatively high inherent stability and can withstand slight changes in conditions, the risk of performing many unnecessary tests can be reduced. Furthermore, when determining the formulation of the preservative solution 17 for the biological drug 11, which has a relatively low inherent stability and is sensitive to even slight changes in conditions, the risk of not performing sufficient tests can be reduced.

[0085] The number of combinations of pH, type and amount of buffer solution 18, and types and amounts of salt, sugar, amino acid, and surfactant is extremely large. It is not easy to select the right combinations to test from this extremely large number of combinations without any guidelines. Therefore, in the technology disclosed herein, a stability range 55 for the first condition 37 is first predicted, and the inherent stability of the biopharmaceutical 11 is determined based on the stability range 55. Then, a search range 22 is derived that leads to the combinations to be tested.

[0086] 18 , the search range derivation unit 48 derives, based on the stable range 55, the search range 22 for testing conditions other than the first condition 37, according to the width of the stable range 55. This reduces the burden on the user U compared to when the user U has to find the search range 22 from the stable range 55.

[0087] As shown in Figure 17, the search range 22 when the stability range 55 is relatively wide is narrower than the search range 22 when the stability range 55 is relatively narrow. A relatively wide stability range 55 corresponds to a case where the inherent stability of the biopharmaceutical 11 is relatively high. Therefore, by narrowing the search range 22, it is possible to avoid performing many unnecessary tests. As a result, the time required for testing can be shortened.

[0088] 17, the search range 22 is narrowed by reducing the variations in the types and / or amounts of the buffer solution 18, salt, sugar, amino acid, and surfactant. This allows for fewer tests that change the types and / or amounts of the buffer solution 18, salt, sugar, amino acid, and surfactant. As a result, the time required for tests that change the types and / or amounts of the buffer solution 18, salt, sugar, amino acid, and surfactant can be shortened.

[0089] 21 , the screen distribution control unit 49 presents the search range 22 to the user U by distributing screen data of the search range display screen 105 to the user terminal 13. Therefore, the user U can easily know the search range 22 without performing complicated analysis.

[0090] As shown in FIG. 6 , the request receiving unit 45 receives a distribution request 15 to obtain sequence information 21 of amino acids constituting the antibody 16 contained in the biopharmaceutical 11. As shown in FIG. 10 , the conversion unit 60 converts the sequence information 21 into vector data 65 representing the characteristics of the antibody 16. As shown in FIG. 11 , the integration unit 61 integrates the vector data 65 and the first condition 37 to generate integrated vector data 66. As shown in FIG. 13 , the first derivation unit 62 derives a stability index value 67 representing the stability of the biopharmaceutical 11 with respect to the first condition 37 from the integrated vector data 66. As shown in FIG. 16 , the second derivation unit 63 derives a stability range 55 based on the stability index value 67. This makes it possible to derive a stability range 55 that is highly reliable and valid.

[0091] The integrating unit 61 and the first derivation unit 62 repeatedly integrate the vector data 65 and the first condition 37 to generate integrated vector data 66 and derive the stability index value 67 from the integrated vector data 66 while changing the first condition 37, thereby obtaining a plurality of stability index values ​​67 corresponding to a plurality of first conditions 37, as shown in Fig. 14. As shown in Fig. 16, the search range derivation unit 48 derives the stability range 55 based on an approximation curve 87 of the plurality of stability index values ​​67. This makes it possible to derive a stability range 55 that is more reliable and more valid.

[0092] 9 and 10 , the conversion unit 60 converts the sequence information 21 into vector data 65 using a conversion model 36 that is a repurposed version of the vectorization unit 71 of a language model 70 that handles the sequence information 21 as a sentence in natural language processing. This makes it possible to easily convert the sequence information 21 into vector data 65 that effectively represents the properties of substances derived from amino acids.

[0093] 11 , the integration unit 61 generates integrated vector data 66 by adding the first condition 37 to all elements of the vector data 65. This makes it possible to obtain integrated vector data 66 that better reflects the first condition 37, and thus the stability index value 67. In other words, the contribution of the first condition 37 to the derivation of the stability index value 67 can be increased.

[0094] 13, the first derivation unit 62 inputs the integrated vector data 66 to the derivation model 38 and causes the derivation model 38 to output a stability index value 67. Therefore, the stability index value 67 with a relatively high accuracy can be easily obtained.

[0095] Biopharmaceuticals 11 containing antibodies 16 as proteins are called antibody drugs and are widely used in the treatment of chronic diseases such as cancer, diabetes, and rheumatoid arthritis, as well as rare diseases such as hemophilia and Crohn's disease. Therefore, this example, in which the substance is a protein and the protein is an antibody 16, can further promote the development of antibody drugs that are widely used in the treatment of various diseases.

[0096] The pH is the most basic condition among the multiple conditions related to the prescription information 20 of the preservative solution 17. Therefore, as shown in Figure 7, if the pH is used as the first condition 37, the original stability of the biopharmaceutical 11 can be determined more accurately.

[0097] Although the integrating unit 61 generates the integrated vector data 66 by adding the first condition 37 to all elements of the vector data 65, this is not limitative. As an example, as shown in Fig. 23 , the integrating unit 61 may generate the integrated vector data 110 by multiplying all elements of the vector data 65 by the first condition 37. This method also makes it possible to obtain the integrated vector data 110, and thus the stability index value 67, that better reflects the first condition 37.

[0098] 24 , the integration unit 61 may add the first condition 37 as a new element N+1 to the vector data 65 to generate the integrated vector data 115. This method makes it easier to generate the integrated vector data 115 than when the first condition 37 is added to or multiplied by all elements of the vector data 65.

[0099] Second Embodiment In the first embodiment, the vector data 65 converted from the sequence information 21 using the conversion model 36 is treated as the first feature, but this is not limiting. As an example, as shown in Fig. 25, an extraction unit 120 may extract a descriptor 121 representing the physical properties of the antibody 16 from the sequence information 21, and the descriptor 121 may be treated as the first feature. Specifically, the descriptor 121 is at least one of FASGAI (Factor Analysis Scales of Generalized Amino Acid Information), Z-scale, Kidera Factors, VHSE (principal components score Vectors of Hydrophobic, Steric, and Electronic properties), Cruciani Properties, and SVP (Subvisible Particles). FASGAI includes multiple elements such as a hydrophobic index, α-helix and β-turn tendencies, bulk properties, compositional property index, flexibility, and electronic properties. The Z-scale also has several components, such as lipophilicity, steric properties (steric bulk and polarizability), electronic properties (polarity and charge), electronegativity, heat of formation, electrophilicity, and hardness.

[0100] By treating the descriptor 121 representing the physical properties of the antibody 16 as the first feature in this way, it is possible to further improve the accuracy of deriving the stability index value 67, and therefore the accuracy of predicting the stability range 55. Note that the descriptor 121 may be treated as the first feature instead of the vector data 65, or the descriptor 121 may be treated as the first feature together with the vector data 65. In the latter case, data obtained by integrating the vector data 65 and the descriptor 121 becomes the first feature.

[0101] Although types 1 to 3 are listed for the buffer solution 18, salt, sugar, amino acid, and surfactant, these are merely examples. The types of buffer solution 18, salt, sugar, amino acid, and surfactant may vary. Furthermore, the number of types may be one or four or more. Similarly, the amounts of buffer solution 18, salt, sugar, amino acid, and surfactant are not limited to the exemplary amounts 1 to 5. The amounts of buffer solution 18, salt, sugar, amino acid, and surfactant may vary. Furthermore, the number of types may be one or six or more. There may also be a search range 22 in which at least one of the conditions of buffer solution 18, salt, sugar, amino acid, and surfactant is not changed. For example, there may also be a search range 22 in which the type and amount of buffer solution 18, salt, amino acid, and surfactant are fixed to one type, and only the type and amount of sugar is changed.

[0102] The first condition 37 is not limited to the hydrogen ion exponent shown in the example, but may be the amount of the buffer solution 18 or the like.

[0103] The method of deriving the search range 22 is not limited to the method of referring to the range correspondence information 39 in the above example. The search range 22 may be derived using a machine learning model. In this case, the vector data 65 and the stable range 55 are input to the machine learning model, and the search range 22 is output from the machine learning model.

[0104] The protein is not limited to the exemplified antibody 16. It may also be a cytokine (interferon, interleukin, etc.), a hormone (insulin, glucagon, follicle-stimulating hormone, erythropoietin, etc.), a growth factor (IGF (insulin-like growth factor)-1, bFGF (basic fibroblast growth factor), etc.), a blood coagulation factor (factor 7, factor 8, factor 9, etc.), an enzyme (lysosomal enzyme, DNA (deoxyribonucleic acid) degrading enzyme, etc.), an Fc (fragment crystalline) fusion protein, a receptor, albumin, or a protein vaccine. The antibody 16 also includes bispecific antibodies, antibody-drug conjugates, low molecular weight antibodies, sugar chain modified antibodies, and the like.

[0105] Furthermore, substances derived from amino acids are not limited to proteins. For example, as shown in Figure 26, they may be peptides 125 or nucleic acids 126. Therefore, macromolecular pharmaceuticals are not limited to biopharmaceuticals 11 that require biotechnology such as genetic recombination technology and cell culture technology, but may also be peptide pharmaceuticals 127 or nucleic acid pharmaceuticals 128 that can be produced using only chemical synthesis technology without requiring biotechnology.

[0106] The search range 22 may not be derived, and the stable range 55 may simply be presented to the user U. Then, the user U may be prompted to devise the search range 22 using the stable range 55 as a clue.

[0107] The drug development support device 12 may be installed in a pharmaceutical manufacturing facility, or may be installed in a data center independent of the pharmaceutical manufacturing facility.

[0108] Instead of distributing screen data of the search range display screen 105 including the search range 22 to the user terminal 13, the search range 22 itself may be distributed to the user terminal 13. In this case, under the control of the browser control unit 97, the user terminal 13 generates the search range display screen 105 based on the search range 22, and displays the search range display screen 105 on the display 29B.

[0109] The method of presenting the search range 22 to the user U is not limited to the example of presenting by distributing screen data. The search range 22 may be presented to the user U by printing it on a paper medium, or by attaching it to an email and sending it to the user terminal 13.

[0110] The hardware configuration of the computer constituting the drug development support device 12 according to the technology of the present disclosure can be modified in various ways. For example, the drug development support device 12 can be configured with multiple computers separated as hardware to improve processing power and reliability. For example, the functions of the request receiving unit 45 and the RW control unit 46, and the functions of the stability range prediction unit 47, the search range derivation unit 48, and the screen distribution control unit 49 can be distributed and performed by two computers. In this case, the drug development support device 12 is configured with two computers. In addition, some or all of the functions of the drug development support device 12 may be performed by the user terminal 13.

[0111] In this way, the hardware configuration of the computer of the pharmaceutical development support device 12 can be changed as appropriate depending on the required performance, such as processing power, safety, and reliability. Furthermore, not only the hardware, but also APs such as the operating program 35 can be duplicated or stored in multiple storage devices in order to ensure safety and reliability.

[0112] In each of the above embodiments, the hardware structure of the processing units that execute various processes, such as the request receiving unit 45, the RW control unit 46, the stable range prediction unit 47, the search range derivation unit 48, the screen delivery control unit 49, the conversion unit 60, the integration unit 61, the first derivation unit 62, the second derivation unit 63, the browser control unit 97, and the extraction unit 120, can be any of the various processors shown below. As described above, the various processors include the CPUs 27A and 27B, which are general-purpose processors that execute software (the operating program 35 and the support AP 95) and function as various processing units, as well as programmable logic devices (PLDs) that are processors whose circuit configuration can be changed after manufacture, such as a field programmable gate array (FPGA), and dedicated electrical circuits that are processors having a circuit configuration designed specifically for executing specific processing, such as an application specific integrated circuit (ASIC).

[0113] A single processing unit may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (e.g., a combination of multiple FPGAs and / or a combination of a CPU and an FPGA).Furthermore, multiple processing units may be configured with a single processor.

[0114] Examples of configuring multiple processing units with a single processor include, first, a form in which one processor is configured with a combination of one or more CPUs and software, as typified by computers such as client and server, and this processor functions as multiple processing units. Second, a form in which a processor is used to realize the functions of the entire system including multiple processing units with a single IC (Integrated Circuit) chip, as typified by systems on chips (SoCs). In this way, various processing units are configured using one or more of the above-mentioned various processors as a hardware structure.

[0115] Furthermore, more specifically, the hardware structure of these various processors can be an electric circuit (circuitry) that combines circuit elements such as semiconductor elements.

[0116] From the above description, the technology described in the following supplementary paragraphs can be understood.

[0117] [Supplementary Item 1] A pharmaceutical development support device comprising a processor, the processor predicting a stability range in which the stability of a macromolecular pharmaceutical will be at or above a set level for a first condition, which is one of a plurality of conditions related to the formulation of a preservation solution of the macromolecular pharmaceutical. [Supplementary Item 2] The pharmaceutical development support device according to Supplementary Item 1, wherein the processor derives a search range for testing conditions other than the first condition based on the stability range, the search range corresponding to the width of the stability range. [Supplementary Item 3] The pharmaceutical development support device according to Supplementary Item 2, wherein the search range is narrower when the stability range is relatively wide than when the stability range is relatively narrow. [Supplementary Item 4] The pharmaceutical development support device according to Supplementary Item 3, wherein the search range is narrowed by reducing at least one of the variations in type and amount of components of the preservation solution. [Supplementary Item 5] The pharmaceutical development support device according to any one of Supplementary Items 2 to 4, wherein the processor presents the search range to a user. [Supplementary Item 6] The pharmaceutical development support device according to any one of Supplementary Items 1 to 5, wherein the processor: acquires sequence information of amino acids constituting the amino acid-derived substance contained in the polymeric drug; converts the sequence information into a first feature representing a characteristic of the substance; integrates the first feature and the first condition to obtain a second feature; derives a stability index value representing the stability of the polymeric drug under the first condition from the second feature; and derives the stability range based on the stability index value. [Supplementary Item 7] The pharmaceutical development support device according to Supplementary Item 6, wherein the processor: acquires a plurality of stability index values ​​corresponding to a plurality of the first conditions by repeating the steps of integrating the first feature and the first condition to obtain the second feature and deriving the stability index value from the second feature while changing the first condition, and deriving the stability index value from the second feature value; and derives the stability range based on an approximation curve of the plurality of stability index values. [Supplementary Item 8] The pharmaceutical development support device according to Supplementary Item 6 or Supplementary Item 7, wherein the processor converts the sequence information into the first feature using a language model that treats the sequence information as a sentence in natural language processing.[Supplementary Item 9] The drug development support device according to any one of Supplementary Items 6 to 8, wherein the processor extracts a descriptor representing a physical property of the substance from the sequence information and treats the descriptor as the first feature. [Supplementary Item 10] The drug development support device according to any one of Supplementary Items 6 to 9, wherein the processor generates the second feature by adding or multiplying the first condition to all elements of the first feature. [Supplementary Item 11] The drug development support device according to any one of Supplementary Items 6 to 9, wherein the processor generates the second feature by adding the first condition as a new element to the first feature. [Supplementary Item 12] The drug development support device according to any one of Supplementary Items 6 to 11, wherein the processor inputs the second feature to a machine learning model and causes the machine learning model to output the stability index value. [Supplementary Item 13] The drug development support device according to any one of Supplementary Items 6 to 12, wherein the substance is any one of a protein, a peptide, and a nucleic acid. [Supplementary Item 14] The pharmaceutical development support device according to Supplementary Item 13, wherein the protein is an antibody. [Supplementary Item 15] The pharmaceutical development support device according to any one of Supplementary Items 1 to 14, wherein the first condition is a hydrogen ion exponent.

[0118] The technology of the present disclosure can be appropriately combined with the various embodiments and / or various modified examples described above. Furthermore, it is not limited to the above embodiments, and various configurations can be adopted without departing from the spirit of the present disclosure. Furthermore, the technology of the present disclosure extends not only to programs, but also to storage media that non-temporarily store programs, and computer program products that include programs.

[0119] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[0120] In this specification, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed by connecting them with "and / or."

[0121] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

Claims

1. A pharmaceutical development support device comprising a processor, the processor predicting a stability range in which the stability of a polymeric pharmaceutical will be at or above a set level for a first condition, which is one of multiple conditions related to the formulation of a preservative solution of the polymeric pharmaceutical.

2. The pharmaceutical development support device according to claim 1, wherein the processor derives a search range for testing conditions other than the first condition based on the stability range, the search range corresponding to the width of the stability range.

3. The pharmaceutical development support device according to claim 2, wherein the search range when the stability range is relatively wide is narrower than the search range when the stability range is relatively narrow.

4. The pharmaceutical development support device according to claim 3, wherein the search range is narrowed by reducing at least one of the variations in type and amount of components of the preservative solution.

5. The pharmaceutical development support device according to claim 2, wherein the processor presents the search range to a user.

6. The pharmaceutical development support device of claim 1, wherein the processor: acquires sequence information of amino acids constituting the amino acid-derived substance contained in the polymeric pharmaceutical; converts the sequence information into a first feature representing a characteristic of the substance; integrates the first feature and the first condition to form a second feature; derives a stability index value representing the stability of the polymeric pharmaceutical against the first condition from the second feature value; and derives the stability range based on the stability index value.

7. The pharmaceutical development support device of claim 6, wherein the processor obtains a plurality of stability index values ​​corresponding to a plurality of first conditions by repeatedly varying the first condition, integrating the first feature and the first condition to obtain the second feature, and deriving the stability index value from the second feature, and deriving the stability range based on an approximation curve of the plurality of stability index values.

8. The pharmaceutical development support device according to claim 6, wherein the processor converts the sequence information into the first feature using a language model that treats the sequence information as a sentence in natural language processing.

9. The pharmaceutical development support device according to claim 6, wherein the processor extracts a descriptor representing the physical properties of the substance from the sequence information, and treats the descriptor as the first feature.

10. A pharmaceutical development support device as described in claim 6, wherein the processor generates the second feature by adding or multiplying the first condition to all elements of the first feature.

11. The pharmaceutical development support device according to claim 6, wherein the processor generates the second feature by adding the first condition as a new element to the first feature.

12. The pharmaceutical development support device according to claim 6, wherein the processor inputs the second feature into a machine learning model and causes the machine learning model to output the stability index value.

13. The pharmaceutical development support device according to claim 6, wherein the substance is one of a protein, a peptide, and a nucleic acid.

14. The pharmaceutical development support device according to claim 13, wherein the protein is an antibody.

15. The pharmaceutical development support device according to claim 1, wherein the first condition is a hydrogen ion exponent.

16. A method for operating a pharmaceutical development support device, comprising: predicting a stability range in which the stability of a polymeric pharmaceutical is at or above a set level for a first condition, which is one of multiple conditions related to the formulation of a preservative solution of the polymeric pharmaceutical.

17. An operating program for a pharmaceutical development support device that causes a computer to execute processing including: predicting a stability range in which the stability of a polymeric pharmaceutical will be at or above a set level for a first condition, which is one of multiple conditions related to the formulation of a preservative solution of the polymeric pharmaceutical.

Images