Drug discovery support device, method for operating drug discovery support device, program for operating drug discovery support device, and evaluation method

Abstract

Provided is a drug discovery support device provided with a processor which acquires measurement data of the extracellular potentials of nerve cells, including local electric field potentials, and selects true synchronous bursts and false synchronous bursts on the basis of the intensity in a frequency band relating to the local electric field potentials.

Description

Drug discovery support device, drug discovery support device operation method, drug discovery support device operation program, and evaluation method

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

[0002] Traditionally, in the drug discovery process, tests have been conducted to evaluate the effects of drug candidate substances on neurons, such as seizure induction. Specifically, before and after administering a candidate substance to neurons derived from human pluripotent stem cells, such as human iPS (induced pluripotent stem) cells, the extracellular potential of the neurons is measured using a microelectrode array (MEA). In this way, measurement data of the extracellular potential before and after administration of the candidate substance is obtained. Then, synchronous bursts are detected in each of the measurement data before and after administration, and the effect of the candidate substance on the neurons is evaluated based on the rate of change in the number of detected synchronous bursts between the measurement data before and after administration. Here, synchronous bursts are a phenomenon that indicates continuous spike firing (bursts) that occurs simultaneously at a high frequency throughout the entire population for a certain period of time in a spontaneously active neuronal population. Synchronized bursts are said to be closely related to the development of neurological disorders involving drug-induced convulsions.

[0003] Japanese Patent No. 6,558,786 and U.S. Patent Specification No. 2019 / 0249147 describe techniques for predicting the effects of a candidate substance on neurons based on various parameters related to synchronized bursts, such as the number of synchronized bursts (NoB: Number of Bursts) and the interval between adjacent synchronized bursts (IBI: Inter Burst Interval).

[0004] Since the detection accuracy is not 100%, there is a possibility that false synchronous bursts (those that are not actually synchronous bursts) may be included among those detected as synchronous bursts. If the detection accuracy of synchronous bursts is low, the reliability of the evaluation results of the effects of candidate substances on neurons will also be low. For this reason, it has been desired to improve the reliability of the evaluation results of the effects of candidate substances on neurons by improving the detection accuracy of synchronous bursts.

[0005] One embodiment of the technology disclosed herein provides a drug discovery support device, an operating method for the drug discovery support device, an operating program for the drug discovery support device, and an evaluation method that can increase the reliability of evaluation results of the effects of candidate pharmaceutical substances on nerve cells more than ever before.

[0006] The drug discovery support device disclosed herein includes a processor, which acquires measurement data of extracellular potentials of neurons, including local field potentials, and selects true and false synchronized bursts based on the intensity in a frequency band related to the local field potentials of the measurement data.

[0007] The processor preferably performs an extraction process to extract data in a frequency band related to local field potentials from the measurement data.

[0008] The processor preferably performs the extraction process using one of the following techniques: a filter, a wavelet transform, and a short-time Fourier transform.

[0009] Preferably, the processor tentatively detects synchronization bursts from the measurement data, and sorts out true synchronization bursts and false synchronization bursts from the tentatively detected synchronization bursts based on the intensity and preset sorting conditions.

[0010] The selection conditions preferably include a threshold value relating to the difference between the intensity and a reference intensity, a threshold value relating to the duration of the change in intensity, and a threshold value relating to the magnitude of the change in intensity.

[0011] When selecting true synchronization bursts and false synchronization bursts from provisionally detected synchronization bursts, it is preferable to use the intensity when no change related to the true synchronization burst occurs as a criterion.

[0012] The processor preferably presents the results of the classification of true and false synchronization bursts to the operator.

[0013] The processor preferably accepts an operator's instruction to correct the sorting results.

[0014] Preferably, the processor uses a trained model generated from training data including waveforms of true synchronization bursts selected based on intensity and waveforms of false synchronization bursts selected based on intensity, inputs a waveform that is unknown as to whether it is a true synchronization burst or a false synchronization burst into the trained model, and outputs a selection result indicating whether it is a true synchronization burst or a false synchronization burst from the trained model.

[0015] The measurement data is preferably measured by a microelectrode array.

[0016] Preferably, the neural cells are derived from human pluripotent stem cells.

[0017] Preferably, the neuronal cells include glutamatergic neuronal cells.

[0018] Preferably, the neurons are co-cultured with astrocytes.

[0019] Preferably, the astrocytes are derived from human pluripotent stem cells.

[0020] The method of operating the drug discovery support device disclosed herein includes acquiring measurement data of the extracellular potential of neurons, which includes local field potentials, and sorting true and false synchronized bursts based on the intensity in a frequency band related to the local field potentials of the measurement data.

[0021] The operating program of the drug discovery support device disclosed herein causes a computer to perform processing including acquiring measurement data of the extracellular potential of neurons, including local field potential, and sorting true and false synchronous bursts based on the intensity in the frequency band related to the local field potential of the measurement data.

[0022] The evaluation method disclosed herein includes acquiring measurement data of extracellular potentials of neurons, including local field potentials, and sorting true synchronized bursts from false synchronized bursts based on the intensity in a frequency band related to the local field potentials of the measurement data.

[0023] It is preferable to evaluate the effect of the test substance on central neurotoxicity.

[0024] According to the technology disclosed herein, it is possible to provide a drug discovery support device, an operating method for the drug discovery support device, an operating program for the drug discovery support device, and an evaluation method that can increase the reliability of the evaluation results of the effects of candidate pharmaceutical substances on nerve cells more than ever before.

[0025] 1 is a diagram illustrating a drug discovery support server, an operator terminal, and a well plate. FIG. 1 is a diagram for explaining true synchronous bursts and false synchronous bursts. FIG. 2 is a block diagram illustrating computers constituting the drug discovery support server and the operator terminal. FIG. 2 is a block diagram illustrating a processing unit of a CPU of the drug discovery support server. FIG. 3 is a diagram illustrating processing of a detection unit. FIG. 4 is a diagram illustrating detection conditions and processing of a detection unit. FIG. 5 is a diagram illustrating detection results. FIG. 6 is a diagram illustrating tentative detection conditions and processing of a tentative detection unit. FIG. 7 is a diagram illustrating tentative detection results. FIG. 8 is a diagram illustrating processing of a selection unit. FIG. 9 is a diagram illustrating processing of a selection unit. FIG. 10 is a diagram illustrating selection conditions and processing of a selection unit. FIG. 11 is a diagram illustrating selection results. FIG. 12 is a block diagram illustrating a processing unit of a CPU of an operator terminal. FIG. 13 is a diagram illustrating a measurement data input screen. FIG. 14 is a diagram illustrating an evaluation result display screen. FIG. 15 is a flowchart illustrating a processing procedure of the drug discovery support server. FIG. 16 is a diagram illustrating another example of processing of the selection unit. FIG. 17 is a diagram illustrating a learning data group. FIG. 18 is a diagram illustrating processing in the learning phase of a selection model. FIG. 19 is a diagram illustrating a state in which a waveform whose identity is unknown as to whether it is a true synchronous burst or a false synchronous burst is input to a selection model, and a selection result indicating whether it is a true synchronous burst or a false synchronous burst is output from the selection model. FIG. 19 is a diagram illustrating a selection result display correction screen.

[0026] First Embodiment As shown in FIG. 1 , a drug discovery support server 10 is connected to an operator terminal 11 via a network 12. The drug discovery support server 10 is an example of a “drug discovery support device” according to the technology of the present disclosure. The operator terminal 11 is installed, for example, at a pharmaceutical company developing drugs or at an organization contracted by a pharmaceutical company to develop drugs, i.e., a contract research organization (CRO). The operator terminal 11 is operated by an operator OP who is involved in drug development at the pharmaceutical company or CRO. The network 12 is, for example, a wide area network (WAN) such as the Internet or a public communication network. Note that while FIG. 1 shows only one operator terminal 11 connected to the drug discovery support server 10, in reality, multiple operator terminals 11 from multiple pharmaceutical companies or CROs are connected to the drug discovery support server 10.

[0027] A well plate 13 is connected to the operator terminal 11. The well plate 13 has a plurality of wells 14. The wells 14 are arranged at equal intervals in the vertical and horizontal directions. FIG. 1 illustrates a well plate 13 having 8 x 6 = 48 wells 14. The wells 14 are numbered. The number of wells 14 is not limited to the illustrated 48, but may be 24, 96, or the like.

[0028] The well 14 is a cylindrical depression with an open top. The well 14 is filled with a culture medium, in which neurons 15 and astrocytes 16 are co-cultured. The neurons 15 are differentiated from human iPS cells 17A, and the astrocytes 16 are differentiated from human iPS cells 17B. In other words, both neurons 15 and astrocytes 16 are derived from human iPS cells 17. The human iPS cells 17 are an example of a "human pluripotent stem cell" according to the technology of the present disclosure. The neurons 15 are glutamatergic neurons. The neurons 15 grow until they exhibit spontaneous activity. Furthermore, a pharmaceutical candidate substance CS is administered to the well 14. The dosage of the candidate substance CS is preset. The candidate substance CS is an example of a "test substance" according to the technology of the present disclosure. Note that the well plate 13 is placed in a thermostatic bath (not shown) to co-culture the neurons 15 and astrocytes 16 under a constant environment.

[0029] A microelectrode array 18 is provided on the bottom surface of the well 14. The microelectrode array 18 is composed of a plurality of electrodes 19 arranged in a square. Fig. 1 shows an example of a microelectrode array 18 having 4 x 4 = 16 electrodes 19. Like the wells 14, the electrodes 19 are also numbered. Note that the number of electrodes 19 is not limited to the 16 shown in the example, and may be 64, 100, or the like.

[0030] The electrodes 19 are exposed on the bottom surface of the well 14 and come into contact with the neurons 15. The electrodes 19 are connected to a potential measurement circuit (not shown). The potential measurement circuit measures the electrical signal from each of the multiple electrodes 19, i.e., the extracellular potential of the neurons 15. Therefore, the extracellular potential is measured for each of the multiple electrodes 19, 16 electrodes 19 in this example. This method of measuring extracellular potential using a microelectrode array 18 is much simpler and less expensive than a method known as manual patch clamp, which measures intracellular potential by inserting electrodes directly into neurons 15. The measurement frequency of the extracellular potential by the potential measurement circuit is, for example, 12.5 kHz. In this case, the extracellular potential is measured 12,500 times per second.

[0031] The measurement results of the extracellular potential from the potential measurement circuit are input into the operator terminal 11. The operator terminal 11 generates measurement data 20, which is a fluctuating waveform that shows the temporal changes in the measurement results of the extracellular potential. Here, the measurement of the extracellular potential is performed for a preset period, for example, 10 minutes. Therefore, the measurement data 20 represents the temporal changes in the measurement results of the extracellular potential for the preset period. In other words, the horizontal axis of the measurement data 20 represents time, and the vertical axis represents the strength (magnitude) of the extracellular potential. The measurement data 20 is stored in association with the number of the well 14 and the number of the electrode 19 where the extracellular potential was measured.

[0032] The frequency band of the measurement data 20 is, for example, 1 Hz to 5 kHz, and therefore the measurement data 20 includes local field potentials (LFPs) in the frequency band of 1 Hz to 50 Hz.

[0033] The operator terminal 11 transmits an evaluation request 21 to the drug discovery support server 10. The evaluation request 21 is a request to have the drug discovery support server 10 evaluate the convulsant potential of the candidate substance CS, which is one of the effects of the candidate substance CS on neurons 15. Convulsant potential is an example of "central neurotoxicity" according to the technology of the present disclosure. The evaluation request 21 includes a measurement data group 22 composed of multiple measurement data 20. The number of measurement data 20 constituting the measurement data group 22 depends on the number of wells 14 and electrodes 19. In this example, there are 48 wells 14, 16 electrodes 19, and two patterns, one before administration and one after administration of the candidate substance CS, so the measurement data group 22 is composed of 48 x 16 x 2 = 1536 measurement data 20. Although not shown in the figure, the evaluation request 21 also includes a terminal ID (Identification Data) for uniquely identifying the operator terminal 11 that is the sender of the evaluation request 21 .

[0034] When the evaluation request 21 is received, the drug discovery support server 10 derives an evaluation result 23 of the convulsant-inducing property of the candidate substance CS. The drug discovery support server 10 distributes the evaluation result 23 to the operator terminal 11 that sent the evaluation request 21. When the evaluation result 23 is received, the operator terminal 11 makes the evaluation result 23 available for viewing by the operator OP.

[0035] The measurement data 20 contains spikes, which are pulses resulting from spontaneous activity of neurons 15. Spikes are also called firings and are responsible for transmitting information between adjacent neurons 15. The phenomenon in which these spikes appear multiple times within a set period is called a burst (continuous firing), and the phenomenon in which bursts are simultaneously observed in the measurement data 20 from electrodes 19 of a set size in one well 14 is called a synchronous burst NB. The drug discovery support server 10 evaluates the convulsion-inducing potential of the candidate substance CS based on this synchronous burst NB.

[0036] As an example, as shown in Fig. 2, the drug discovery support server 10 provisionally detects a synchronization burst NB from the measurement data 20. Fig. 2 illustrates a case where two synchronization bursts NB are provisionally detected.

[0037] Among those provisionally detected as synchronization bursts NB, there is a possibility that false synchronization bursts FNB (those that are not actually synchronization bursts NB) are included. For this reason, it is necessary to sort out whether the provisionally detected synchronization bursts NB are true synchronization bursts TNB or false synchronization bursts FNB.

[0038] As shown in graph 25, when a true synchronization burst TNB occurs, it is known that a change in intensity is observed in the frequency band (1 Hz to 50 Hz) related to the LFP. Therefore, the drug discovery support server 10 distinguishes whether the provisionally detected synchronization burst NB is a true synchronization burst TNB or a false synchronization burst FNB based on the presence or absence of a change in intensity in the frequency band related to the LFP.

[0039] 3, the computers that make up the drug discovery support server 10 and the operator terminal 11 basically have the same configuration, and include a storage 30, a memory 31, a CPU (Central Processing Unit) 32, a communication unit 33, a display 34, and an input device 35. These are interconnected via a bus line 36.

[0040] The storage 30 is a hard disk drive built into the computer that constitutes the drug discovery support server 10 and the operator terminal 11, or connected via a cable or network. Alternatively, the storage 30 is a disk array with multiple hard disk drives connected in series. The storage 30 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.

[0041] The memory 31 is a work memory for the CPU 32 to execute processing. The CPU 32 loads programs stored in the storage 30 into the memory 31 and executes processing in accordance with the programs. In this way, the CPU 32 comprehensively controls each part of the computer. The CPU 32 is an example of a "processor" according to the technology of the present disclosure. The memory 31 may be built into the CPU 32.

[0042] The communication unit 33 is a network interface that controls the transmission of various information via the network 12, etc. The display 34 displays various screens. The various screens are equipped with operation functions using a GUI (Graphical User Interface). The computers that make up the drug discovery support server 10 and the operator terminal 11 accept input of operation instructions from an input device 35 via the various screens. The input device 35 is a keyboard, a mouse, a touch panel, a microphone for voice input, etc.

[0043] In the following explanation, the parts of the computer that make up the drug discovery support server 10 (storage 30 and CPU 32) are distinguished by adding the suffix "A" to their symbols, and the parts of the computer that make up the operator terminal 11 (storage 30, CPU 32, display 34, and input device 35) are distinguished by adding the suffix "B" to their symbols.

[0044] As an example, as shown in FIG. 4 , an operating program 40 is stored in the storage 30A of the drug discovery support server 10. The operating program 40 is an AP for causing a computer to function as the drug discovery support server 10. In other words, the operating program 40 is an example of an "operating program for a drug discovery support device" according to the technology of the present disclosure. The storage 30 also stores detection conditions 41, tentative detection conditions 42, and selection conditions 43. Although not shown to avoid complexity, the storage 30 also stores a high-pass filter 58 (see FIG. 5 ) and a low-pass filter 63 (see FIG. 10 ).

[0045] When the operating program 40 is started, the CPU 32A of the computer constituting the drug discovery support server 10 works in cooperation with the memory 31, etc. to function as a request receiving unit 45, a read / write (hereinafter abbreviated as RW (Read Write)) control unit 46, a detection unit 47, a tentative detection unit 48, a selection unit 49, an evaluation unit 50, and a screen distribution control unit 51.

[0046] The request receiving unit 45 receives various requests from the operator terminal 11, including the evaluation request 21. As described above, the evaluation request 21 includes a measurement data group 22 made up of multiple pieces of measurement data 20. Therefore, by receiving the evaluation request 21, the request receiving unit 45 acquires the measurement data 20. When the evaluation request 21 is received, the request receiving unit 45 outputs the measurement data group 22 included in the evaluation request 21 to the RW control unit 46. Furthermore, although not shown in the figure, the request receiving unit 45 outputs the terminal ID of the operator terminal 11 included in the evaluation request 21 to the screen distribution control unit 51.

[0047] The RW control unit 46 controls the storage of various data in the storage 30A and the reading of various data from the storage 30A. In particular, the RW control unit 46 controls the storage of the measurement data group 22 in the storage 30A and the reading of the measurement data group 22 from the storage 30A. The RW control unit 46 outputs the read measurement data group 22 to the detection unit 47 and the selection unit 49.

[0048] The RW control unit 46 reads out the detection conditions 41 from the storage 30A and outputs the read out detection conditions 41 to the detection unit 47. The RW control unit 46 also reads out the tentative detection conditions 42 from the storage 30A and outputs the read out tentative detection conditions 42 to the tentative detection unit 48. The RW control unit 46 also reads out the sorting conditions 43 from the storage 30A and outputs the read out sorting conditions 43 to the sorting unit 49.

[0049] The RW control unit 46 reads the high-pass filter 58 from the storage 30A and outputs the read high-pass filter 58 to the detection unit 47. The RW control unit 46 also reads the low-pass filter 63 from the storage 30A and outputs the read low-pass filter 63 to the selection unit 49.

[0050] The detecting unit 47 detects spikes from the measurement data 20 based on the detection conditions 41. The detecting unit 47 outputs spike detection results 55 to the tentative detecting unit 48 and the selecting unit 49.

[0051] The tentative detection unit 48 tentatively detects the synchronization burst NB based on the tentative detection condition 42. The tentative detection unit 48 outputs a tentative detection result 56 of the synchronization burst NB to the selection unit 49.

[0052] The selection unit 49 selects true synchronization bursts TNB and false synchronization bursts FNB from the provisionally detected synchronization bursts NB based on the selection conditions 43. The selection unit 49 outputs a selection result 57 of true synchronization bursts TNB and false synchronization bursts FNB to the evaluation unit 50.

[0053] The evaluation unit 50 evaluates the convulsant-inducing activity of the candidate substance CS based on the selection result 57. The evaluation unit 50 outputs the evaluation result 23 of the convulsant-inducing activity of the candidate substance CS to the screen distribution control unit 51.

[0054] The screen delivery control unit 51 controls the delivery of various screens to the operator terminal 11. Specifically, the screen delivery control unit 51 delivers and outputs various screens to the operator terminal 11 that has sent the various requests in the form of screen data for web delivery created using a markup language such as XML (Extensible Markup Language). At this time, the screen delivery control unit 51 identifies the operator terminal 11 that has sent the various requests based on the terminal ID from the request receiving unit 45. Note that instead of XML, other data description languages ​​such as JSON (Javascript (registered trademark) Object Notation) may be used.

[0055] The various screens include a measurement data input screen 75 (see FIG. 16) for inputting measurement data 20, and an evaluation result display screen 80 (see FIG. 17) for presenting evaluation results 23 to an operator OP. In addition to these processing units 45 to 51, the CPU 32A also includes an instruction receiving unit 115 (see FIG. 23) for receiving various operation instructions from the input device 35.

[0056] As an example, as shown in FIG. 5 , the detection unit 47 applies a high-pass filter 58 to the measurement data 20 to derive high-frequency measurement data 20H from the measurement data 20. The high-pass filter 58 passes data in the frequency band of 200 Hz to 5 kHz. In other words, the high-pass filter 58 blocks data in the frequency band below 200 Hz. Therefore, the high-frequency measurement data 20H has a frequency band of 200 Hz to 5 kHz. Data in the relatively high-frequency band of 200 Hz to 5 kHz is sufficient for spike detection; data in the relatively low-frequency band, such as LFPs of 1 Hz to 50 Hz, constitutes noise. Therefore, the detection unit 47 derives the high-frequency measurement data 20H using the high-pass filter 58. Instead of the high-pass filter 58, a band-pass filter that passes only data in a frequency band with a relatively good signal-to-noise ratio may be used to reduce noise and improve spike detection accuracy. A frequency band with a relatively good S / N ratio is, for example, a frequency band from 200 Hz to 3 kHz.

[0057] 6, the detection condition 41 is that the absolute value of the intensity of the extracellular potential is equal to or greater than the threshold intensity THI. Therefore, the detector 47 detects, as spikes, waveforms in the high-frequency measurement data 20H whose absolute value of the intensity is equal to or greater than the threshold intensity THI.

[0058] The threshold strength THI is based on, for example, the strength of the baseline when no change in strength associated with a true synchronous burst occurs. More specifically, the threshold strength THI is an integer multiple (e.g., six times) of the representative value of the baseline strength when no change in strength associated with a true synchronous burst occurs. The representative value may be the mean, median, or mode of the baseline strength itself, or the mean, median, or mode of the standard deviation of the baseline strength per unit time (e.g., per second). The threshold strength THI is updated at set intervals (e.g., every 10 seconds). The baseline when no change in strength associated with a true synchronous burst TNB occurs is selected by the operator OP. The same applies to the subsequent thresholds.

[0059] 7, the detection result 55 is information in which the spike detection times for each electrode 19 in each well 14 before and after administration of the candidate substance CS are registered. The detection times are values ​​measured with the measurement start time of the measurement data 20 set to 0.

[0060] 8 , the provisional detection unit 48 generates a spike histogram 60 and a raster plot 61 based on the detection results 55. The spike histogram 60 is a graph showing the number of spikes detected per unit time (e.g., one second) in one well 14, with the horizontal axis representing time and the vertical axis representing the number of spikes detected. The raster plot 61 is a graph in which strip-shaped markers are added to the locations where spikes are detected for each electrode 19 in one well 14, with the horizontal axis representing time.

[0061] The tentative detection condition 42 includes the following three conditions: 1. The number of spike detections is equal to or greater than the threshold detection number THND. 2. A burst occurs at an electrode 19 that is equal to or greater than the threshold rate THP1. 3. If the interval between two adjacent waveforms that satisfy conditions 1 and 2 is less than the threshold interval THIV, one of the waveforms is adopted.

[0062] Therefore, the tentative detection unit 48 detects waveforms whose spike detection count is equal to or greater than the threshold detection count THND from the spike histogram 60. The tentative detection unit 48 also derives the burst occurrence rate during the occurrence time of waveforms whose spike detection count is equal to or greater than the threshold detection count THND from the raster plot 61. Then, it detects waveforms whose derived occurrence rate is equal to or greater than the threshold rate THP1. Finally, if the interval between two adjacent waveforms whose spike detection count is equal to or greater than the threshold detection count THND and whose burst occurrence rate is equal to or greater than the threshold rate THP1 is less than the threshold interval THIV, the tentative detection unit 48 adopts one of the waveforms. For example, the waveform with the later detection time of the two waveforms may be adopted. The waveform with the greater spike detection count or the greater occurrence rate of the two waveforms may also be adopted. In this way, the tentative detection unit 48 tentatively detects a waveform that satisfies the tentative detection condition 42 as a synchronous burst NB.

[0063] The threshold detection number THND is based on the number of spike detections in the baseline when no change in intensity related to a true synchronous burst TNB occurs. More specifically, the threshold detection number THND is an integer multiple (e.g., six times) of the representative value of the number of spike detections in the baseline when no change in intensity related to a true synchronous burst TNB occurs. The representative value may be the mean, median, or mode of the number of spike detections in the baseline itself, or the mean, median, or mode of the standard deviation of the number of spike detections in the baseline per unit time (e.g., per second). The threshold detection number THND is updated at set intervals (e.g., every 10 seconds).

[0064] The threshold ratio THP1 is, for example, 75%. In this example, since the number of electrodes 19 is 16, the threshold ratio THP1 is 16×0.75=12.

[0065] The threshold interval THIV is a representative value of the interval between two adjacent waveforms that satisfy conditions 1 and 2. The representative value is the minimum value of the interval between two adjacent waveforms that satisfy conditions 1 and 2 (waveforms that are determined to be separate synchronization bursts NB).

[0066] 9, the provisional detection result 56 is information in which the provisional detection time of the synchronous burst NB is registered for each well 14 before and after administration of the candidate substance CS. Like the spike detection time, the provisional detection time is a value measured with the measurement start time of the measurement data 20 set to 0.

[0067] As an example, as shown in FIG. 10 , the selection unit 49 applies a low-pass filter 63 to the measurement data 20 to derive low-frequency measurement data 20L from the measurement data 20. The low-pass filter 63 passes data in the frequency band of 1 Hz to 50 Hz, which is the frequency band associated with LFP, out of the frequency band of 1 Hz to 5 kHz of the measurement data 20. In other words, the low-pass filter 63 blocks data in frequency bands higher than 50 Hz from passing through. Therefore, the low-frequency measurement data 20L has a frequency band of 1 Hz to 50 Hz. The low-pass filter 63 is an example of a "filter" according to the technology disclosed herein. The low-frequency measurement data 20L is an example of "data in a frequency band associated with local field potentials" according to the technology disclosed herein. The process of applying the low-pass filter 63 to the measurement data 20 to derive the low-frequency measurement data 20L is an example of an "extraction process" according to the technology disclosed herein. It should be noted that a bandpass filter that passes only data in a frequency band necessary for distinguishing between true synchronization bursts TNB and false synchronization bursts FNB may be used instead of the lowpass filter 63. The frequency band necessary for distinguishing between true synchronization bursts TNB and false synchronization bursts FNB is, for example, a frequency band of 2 Hz to 10 Hz.

[0068] As an example, as shown in FIG. 11 , the selection unit 49 calculates the power spectral density (PSD) for the low-frequency measurement data 20L and generates a PSD graph 65. The power spectral density PSD is a value obtained by dividing the square of the intensity (power) by the frequency resolution. The horizontal axis of the PSD graph 65 represents time, and the vertical axis represents the power spectral density PSD. The power spectral density PSD is an example of the "intensity in a frequency band related to the local field potential" according to the technology of the present disclosure.

[0069] 12, the selection condition 43 includes the following three conditions: 1. The power spectrum density PSD is equal to or greater than a threshold density THD; 2. The state 1 continues for a threshold time THT or more; and 3. The conditions 1 and 2 are satisfied by an electrode 19 having a threshold rate THP2 or more.

[0070] The selection unit 49 generates a raster plot 66 of the power spectral density PSD. The raster plot 66 of the power spectral density PSD is a graph in which, for each electrode 19 in one well 14, strip markers are added to the locations that satisfy the selection conditions 1 and 2 of the selection condition 43, and the horizontal axis represents time.

[0071] The selection unit 49 detects waveforms from the PSD graph 65 in which the power spectral density PSD is equal to or greater than the threshold density THD and in which the state in which the power spectral density PSD is equal to or greater than the threshold density THD continues for a threshold time THT or longer. The selection unit 49 also derives, from the power spectral density PSD raster plot 66, the occurrence rate of waveforms in which the power spectral density PSD is equal to or greater than the threshold density THD and in which the state in which the power spectral density PSD is equal to or greater than the threshold density THD continues for a threshold time THT or longer. The selection unit 49 then selects waveforms whose derived occurrence rate is equal to or greater than the threshold rate THP2 as true synchronization bursts TNB. On the other hand, the selection unit 49 selects waveforms that do not satisfy any one of the conditions 1 to 3 of the selection condition 43 as false synchronization bursts FNB.

[0072] The threshold density THD is based on the power spectral density PSD of the baseline when no change in intensity related to a true synchronization burst TNB occurs. More specifically, the threshold density THD is an integer multiple (e.g., six times) of the representative value of the power spectral density PSD of the baseline when no change in intensity related to a true synchronization burst TNB occurs. The representative value is the mean, median, mode, etc. of the baseline power spectral density PSD itself. The threshold density THD is updated at set intervals (e.g., every 10 seconds).

[0073] The threshold time THT is, for example, 5 to 10 seconds. The threshold percentage THP2 is, for example, 75%, the same as the threshold percentage THP1. Therefore, in this example, the threshold percentage THP2 is also 12, the same as the threshold percentage THP1. The threshold density THD is an example of a "threshold related to the difference from the intensity reference" according to the technology of the present disclosure. The threshold time THT is an example of a "threshold related to the duration of the change in intensity" according to the technology of the present disclosure. Furthermore, the threshold percentage THP2 is an example of a "threshold related to the magnitude of the occurrence of a change in intensity" according to the technology of the present disclosure.

[0074] As an example, as shown in FIG. 13, the selection result 57 is information in which each provisionally detected synchronization burst in the provisional detection result 56 is registered as to whether it is a true synchronization burst TNB or a false synchronization burst FNB.

[0075] The evaluation unit 50 refers to the selection result 57 and counts the number of detected true synchronous bursts TNB before and after administration of the candidate substance CS for each well 14. As shown in FIG. 14 as an example, the evaluation unit 50 subtracts the number of detected true synchronous bursts TNB before administration of the candidate substance CS from the number of detected true synchronous bursts TNB after administration of the candidate substance CS, divides the subtraction result by the number of detected true synchronous bursts TNB before administration of the candidate substance CS, and multiplies the result by 100 to calculate the rate of change in synchronous burst occurrence. The evaluation unit 50 calculates the rate of change in synchronous burst occurrence for each well 14 and outputs a representative value, such as the average value, as the evaluation result 23. The evaluation unit 50 also outputs the result of determining whether or not the candidate substance CS has convulsive inducing properties based on the rate of change in synchronous burst occurrence as the evaluation result 23. For example, the evaluation unit 50 determines that the candidate substance CS does not have convulsive inducing properties when the rate of change in synchronous burst occurrence is less than 30%, and determines that the candidate substance CS has convulsive inducing properties when the rate of change in synchronous burst occurrence is 30% or more. Figure 14 illustrates an example where the rate of change in synchronous burst occurrence is 200%, indicating that the candidate substance CS has convulsive inducing properties. The threshold value of the rate of change in synchronous burst occurrence used to determine the presence or absence of convulsive inducing properties is not limited to the exemplary 30%, but may be 50%, 100%, etc. Furthermore, the threshold value of the rate of change in synchronous burst occurrence used to determine the presence or absence of convulsive inducing properties may be configured to be changeable by the operator OP.

[0076] As an example, as shown in FIG. 15 , an evaluation AP 70 is stored in the storage 30B of the operator terminal 11. The evaluation AP 70 is installed in the operator terminal 11 by the operator OP. The evaluation AP 70 is an AP that causes the drug discovery support server 10 to evaluate the convulsant-inducing potential of a candidate substance CS. When the evaluation AP 70 is activated, the CPU 32B of the operator terminal 11 functions as a browser control unit 72 in cooperation with the memory 31 and the like. The browser control unit 72 controls the operation of a web browser dedicated to the evaluation AP 70.

[0077] The browser control unit 72 reproduces various screens based on various screen data from the drug discovery support server 10 and displays the reproduced various screens on the display 34B. The browser control unit 72 also accepts various operation instructions input by the operator OP from the input device 35B via the various screens. The browser control unit 72 transmits various requests, including the evaluation request 21, to the drug discovery support server 10 in response to the operation instructions.

[0078] When the evaluation AP 70 is launched, a measurement data input screen 75, as shown in FIG. 16 as an example, is displayed on the display 34B under the control of the browser control unit 72. The measurement data input screen 75 has an input box 76 for the measurement data group 22. A file of the measurement data group 22 can be dropped into the input box 76. After dropping the file of the desired measurement data group 22 into the input box 76, the operator OP selects the evaluation button 77. When the evaluation button 77 is selected, the browser control unit 72 generates an evaluation request 21 including the measurement data group 22 input into the input box 76, and transmits the generated evaluation request 21 to the drug discovery support server 10.

[0079] Furthermore, when the convulsant-inducing potential of a candidate substance is evaluated in the drug discovery support server 10, an evaluation result display screen 80, as shown in FIG. 17 as an example, is displayed on the display 34B under the control of the browser control unit 72. The evaluation result display screen 80 displays the evaluation result 23, which is the rate of change in synchronous burst occurrence and the determination result of the presence or absence of convulsant-inducing potential. In this way, the evaluation result 23 is presented to the operator OP in the form of distributed screen data.

[0080] A save button 81 and an OK button 82 are provided at the bottom of the evaluation result display screen 80. When the save button 81 is selected, the display contents of the evaluation result display screen 80, including the evaluation results 23, are stored in the storage 30B of the operator terminal 11. When the OK button 82 is selected, the display of the evaluation result display screen 80 is cleared.

[0081] Next, the operation of the above configuration will be described with reference to the flowchart shown in Fig. 18 as an example. When the operating program 40 is started in the drug discovery support server 10, the CPU 32A of the drug discovery support server 10 functions as a request receiving unit 45, an RW control unit 46, a detection unit 47, a tentative detection unit 48, a selection unit 49, an evaluation unit 50, and a screen distribution control unit 51, as shown in Fig. 4. When the evaluation AP 70 is started in the operator terminal 11, the CPU 32B of the operator terminal 11 functions as a browser control unit 72, as shown in Fig. 15.

[0082] 16 is displayed on the display 34B of the operator terminal 11 under the control of the browser control unit 72. When the operator OP inputs a file of the desired measurement data group 22 into the input box 76 on the measurement data input screen 75 and selects the evaluation button 77, an evaluation request 21 is sent from the browser control unit 72 to the drug discovery support server 10.

[0083] In the drug discovery support server 10, the request receiving unit 45 receives the evaluation request 21 (YES in step ST100). As a result, the request receiving unit 45 acquires the measurement data group 22 included in the evaluation request 21. The measurement data group 22 is output from the request receiving unit 45 to the RW control unit 46 and stored in the storage 30A under the control of the RW control unit 46 (step ST110). In addition, the terminal ID of the operator terminal 11 included in the evaluation request 21 is output from the request receiving unit 45 to the screen distribution control unit 51.

[0084] The measurement data group 22 is read from the storage 30A by the RW control unit 46 (step ST120). The measurement data group 22 is output from the RW control unit 46 to the detection unit 47 and the selection unit 49.

[0085] As shown in Fig. 5, in the detection unit 47, a high-pass filter 58 is applied to the measurement data 20, thereby deriving high-frequency measurement data 20H. Next, as shown in Fig. 6, spikes are detected from the measurement data 20 based on the detection conditions 41, and the detection result 55 shown in Fig. 7 is generated (step ST130). The detection result 55 is output from the detection unit 47 to the tentative detection unit 48 and the selection unit 49.

[0086] 8, the synchronization burst NB is tentatively detected in the tentative detection unit 48, and the tentative detection result 56 shown in Fig. 9 is generated (step ST140). The tentative detection result 56 is output from the tentative detection unit 48 to the selection unit 49.

[0087] As shown in Figures 10 to 12, the selection unit 49 selects true synchronization bursts TNB and false synchronization bursts FNB from the provisionally detected synchronization bursts NB (step ST150). More specifically, first, as shown in Figure 10, a low-pass filter 63 is applied to the measurement data 20, thereby deriving low-frequency measurement data 20L. Next, as shown in Figure 11, the power spectral density PSD of the intensity of the low-frequency measurement data 20L is calculated. Finally, as shown in Figure 12, true synchronization bursts TNB and false synchronization bursts FNB are selected based on the selection condition 43, and the selection result 57 shown in Figure 13 is generated. The selection result 57 is output from the selection unit 49 to the evaluation unit 50.

[0088] 14, the evaluation unit 50 calculates the rate of change in synchronous burst occurrence from the number of detected true synchronous bursts in the selection result 57, and determines whether the candidate substance CS has convulsant-inducing properties based on the rate of change in synchronous burst occurrence (step ST160). The rate of change in synchronous burst occurrence and the determination results of whether the candidate substance CS has convulsant-inducing properties are output from the evaluation unit 50 to the screen distribution control unit 51 as the evaluation result 23.

[0089] The screen distribution control unit 51 generates screen data for the evaluation result display screen 80 shown in Fig. 17 based on the evaluation result 23. Under the control of the screen distribution control unit 51, the screen data for the evaluation result display screen 80 is distributed to the operator terminal 11 that is the sender of the evaluation request 21 (step ST170).

[0090] In the operator terminal 11, under the control of the browser control unit 72, the screen data of the evaluation result display screen 80 is reproduced, and the reproduced evaluation result display screen 80 is displayed on the display 34B. In this way, the evaluation result 23 is presented to the operator OP.

[0091] As described above, the request receiving unit 45 of the drug discovery support server 10 receives an evaluation request 21 to acquire measurement data 20 of extracellular potentials of neurons 15, including LFPs. The selection unit 49 calculates a power spectral density (PSD) as the intensity in the frequency band associated with the LFPs of the measurement data 20. The selection unit 49 then selects true synchronized bursts (TNB) and false synchronized bursts (FNB) based on the power spectral density (PSD). This reduces the likelihood that false synchronized bursts (FNB) are included among those ultimately detected as synchronized bursts (NB). This makes it possible to improve the reliability of the evaluation results 23 of the effects of a candidate substance (CS) on neurons 15, such as its convulsant-inducing potential, compared to conventional methods.

[0092] 10, the selection unit 49 performs an extraction process to extract low-frequency measurement data 20L, which is data in the frequency band related to the LFP, from the measurement data 20. This allows accurate selection between true synchronization bursts TNB and false synchronization bursts FNB based on the intensity in the frequency band related to the LFP.

[0093] 10, the selection unit 49 performs the extraction process using a low-pass filter 63. Therefore, it is possible to easily extract the low-frequency measurement data 20L, which is data in the frequency band related to the LFP.

[0094] As shown in Fig. 8, the tentative detection unit 48 tentatively detects synchronization bursts NB from the measurement data 20. Also, as shown in Fig. 12, the selection unit 49 selects true synchronization bursts TNB and false synchronization bursts FNB from the tentatively detected synchronization bursts NB based on the power spectral density PSD and preset selection conditions 43. This eliminates the need for the operator OP to tentatively detect synchronization bursts NB and to select true synchronization bursts TNB and false synchronization bursts FNB.

[0095] 12, the selection condition 43 includes a threshold density THD relating to the difference between the power spectral density PSD and a reference, a threshold time THT relating to the duration of a change in the power spectral density PSD, and a threshold percentage THP2 relating to the magnitude of the change in the power spectral density PSD. This makes it possible to easily and reliably select true synchronization bursts TNB from false synchronization bursts FNB.

[0096] 12, when selecting a true synchronization burst TNB and a false synchronization burst FNB from the provisionally detected synchronization burst NB, the power spectral density PSD when no change in intensity related to the true synchronization burst TNB occurs is used as a reference, thereby improving the reliability of the selection result 57 of the true synchronization burst TNB and the false synchronization burst FNB.

[0097] 1, the measurement data 20 is measured by the microelectrode array 18. Therefore, spikes and synchronous bursts can be detected from the measurement data 20, and the convulsion-inducing ability of the candidate substance CS can be evaluated based on the measurement data 20.

[0098] As shown in Figure 1, the neuron 15 is derived from the human iPS cell 17A, which allows accurate evaluation of the convulsant-inducing potential of the candidate substance CS in humans.

[0099] As shown in Figure 1, the neuron 15 is a glutamatergic neuron. Glutamatergic neurons are abundant in the cerebral cortex, which is deeply involved in the occurrence of seizures, and are responsible for excitatory transmission. Therefore, they are suitable for evaluating the convulsion-inducing properties of the candidate substance CS.

[0100] As shown in Figure 1, neurons 15 are co-cultured with astrocytes 16. By co-culturing neurons 15 and astrocytes 16, which cooperate in the human body, it is possible to reproduce the interaction between neurons 15 and astrocytes 16 in the human body. Note that instead of or in addition to astrocytes 16, microglia and neurons 15 may be co-cultured.

[0101] As shown in Figure 1, the astrocytes 16 were derived from human iPS cells 17B, which allows accurate evaluation of the convulsant-inducing potential of the candidate substance CS in humans.

[0102] Although an example using a low-pass filter 63 for extraction processing has been given, this is not limiting. As an example, as shown in FIG. 19 , a wavelet transform or a short-time Fourier transform may be applied to the measurement data 20 as extraction processing. In this case, the selection unit 49 outputs the results of the wavelet transform or short-time Fourier transform as an amplitude scalogram 85 showing the time-series changes in power for each frequency band. The amplitude scalogram 85 represents the magnitude of power using different colors, with the horizontal axis representing time and the vertical axis representing logarithmic frequency. The selection unit 49 selects true synchronization bursts (TNB) and false synchronization bursts (FNB) based on the power in the frequency band related to the LFP in the amplitude scalogram 85 (e.g., the portion indicated by the dashed rectangular frame and arrow). In this case, power is an example of the "intensity in the frequency band related to the local field potential" according to the technology of the present disclosure.

[0103] In this way, by performing extraction processing using the wavelet transform or short-time Fourier transform technique, it is possible to easily extract power in the frequency band related to LFP.

[0104] [Second Embodiment] As in the first embodiment, by repeatedly selecting true synchronization bursts TNB and false synchronization bursts FNB based on the intensity in the frequency band related to the LFP, a set of waveforms 91 (see FIG. 20) including synchronization bursts NB and the selection results 57 is accumulated as data. Therefore, in this embodiment, the set of waveforms 91 and the selection results 57 accumulated as data is effectively utilized.

[0105] 20 as an example, in this embodiment, a set of a waveform 91 stored as data and a selection result 57 is used as a set of a learning waveform 91L and a correct selection result 57CA, which are used as learning data 100. A plurality of learning data 100 are collected to form a learning data group 101. The learning waveform 91L includes a waveform of a true synchronization burst TNB selected based on the intensity in a frequency band related to the provisionally detected synchronization burst NB and the LFP, and a waveform of a false synchronization burst FNB selected based on the intensity in a frequency band related to the provisionally detected synchronization burst NB and the LFP.

[0106] As an example, as shown in FIG. 21 , training data 100 is used to train a selection model 105. That is, a training waveform 91L is input to the selection model 105, which then outputs a training selection result 57L. The training selection result 57L is compared with a correct answer selection result 57CA, and a loss calculation for the selection model 105 is performed using a loss function based on the comparison result. Then, internal parameters such as filter coefficients of the selection model 105 are updated according to the result of the loss calculation, and the selection model 105 is updated according to the updated settings. Note that the selection model 105 is a machine learning model constructed using a machine learning technique such as a neural network, a support vector machine, a gradient boosting tree, Adaboost, or a random forest. The selection model 105 is an example of a “trained model” according to the technology disclosed herein.

[0107] The above-described series of processes, including inputting the learning waveform 91L to the selection model 105, outputting the learning selection result 57L from the selection model 105, calculating the loss, setting the update, and updating the selection model 105, are repeatedly performed while the training data 100 is changed. The repetition of the above-described series of processes is terminated when the selection accuracy of the learning selection result 57L relative to the correct selection result 57CA reaches a preset level. The selection model 105 whose selection accuracy has reached the preset level is stored in the storage 30A of the drug discovery support server 10 and used by the selection unit 49. Regardless of the selection accuracy, learning may be terminated when the above-described series of processes have been repeated a predetermined number of times. Furthermore, learning of the selection model 105 may continue even after storage in the storage 30A. Furthermore, learning of the selection model 105 may be performed by the drug discovery support server 10 or by a device separate from the drug discovery support server 10. In the latter case, the selection model 105 is transmitted to the drug discovery support server 10 from a separate device.

[0108] 22 as an example, in this embodiment, the selection unit 49 inputs a waveform 91 including a provisionally detected synchronization burst NB, and it is unknown whether the waveform 91 is a true synchronization burst TNB or a false synchronization burst FNB, to the selection model 105. Then, the selection model 105 outputs a selection result 57 indicating whether the waveform 91 is a true synchronization burst TNB or a false synchronization burst FNB. FIG. 22 illustrates a case where the provisionally detected synchronization burst NB of the waveform 91 is selected as a false synchronization burst FNB.

[0109] As described above, in this embodiment, the selection unit 49 uses a selection model 105 generated from training data 100 including a waveform 91 of a true synchronization burst TNB selected based on the intensity in a frequency band associated with an LFP from a provisionally detected synchronization burst NB, and a waveform 91 of a false synchronization burst FNB selected based on the intensity in a frequency band associated with an LFP from a provisionally detected synchronization burst NB. The selection unit 49 inputs the waveform 91, which is unknown as to whether it is a true synchronization burst TNB or a false synchronization burst FNB, to the selection model 105, and causes the selection model 105 to output a selection result 57 indicating whether the waveform 91 is a true synchronization burst TNB or a false synchronization burst FNB. Therefore, the selection result 57 can be easily obtained using the selection model 105 generated by effectively utilizing the set of the waveform 91 and the selection result 57 stored as data. This eliminates the need to calculate a power spectral density PSD and compare it with the selection condition 43.

[0110] The following method can also be considered as a method for utilizing the set of waveform 91 and selection result 57 stored as data. That is, the threshold strength THI of detection condition 41, the threshold detection count THND, threshold percentage THP1, and threshold interval THIV of tentative detection condition 42, and the threshold density THD, threshold time THT, and threshold percentage THP2 of selection condition 43 are variously changed, and each time, for waveform 91 for which selection result 57 is known, a selection result 57 based on the intensity in the frequency band related to LFP is output. Then, the degree of agreement (accuracy rate) between the output selection result 57 and the known selection result 57 is calculated, and the combination of thresholds that results in the highest degree of agreement is derived. A well-known method for solving an optimization problem can be used to derive the combination of thresholds that results in the highest degree of agreement.

[0111] [Third Embodiment] In this embodiment, the screen delivery control unit 51 receives the selection result 57 from the selection unit 49. Then, based on the selection result 57, a selection result display correction screen 110 shown in Fig. 23 as an example is generated and delivered to the operator terminal 11. The selection result display correction screen 110 displays the waveforms 91 extracted by the tentative detection unit 48 side by side.

[0112] Below the waveform 91, a selection result 57, a circle button 111, and an x ​​button 112 are provided. The operator OP, for example, observes the waveform 91 to select whether the synchronization burst NB in ​​the waveform 91 is a true synchronization burst TNB or a false synchronization burst FNB. If the selection result 57 is the same as the result selected by the operator OP, the operator OP considers the selection result 57 to be correct and selects the circle button 111. On the other hand, if the selection result 57 is different from the result selected by the operator OP, the operator OP considers the selection result 57 to be incorrect and selects the x button 112. The operator OP selects either the circle button 111 or the x button 112 for all waveforms 91 displayed on the selection result display correction screen 110, and then selects the OK button 113. When the OK button 113 is selected, a correction instruction corresponding to the selection state of the circle button 111 or the x button 112 is received by the instruction receiving unit 115.

[0113] When a correction instruction is received, the instruction receiving unit 115 outputs the correction instruction to the evaluation unit 50. The evaluation unit 50 corrects the selection result 57 from the selection unit 49 in accordance with the correction instruction. The evaluation unit 50 outputs the evaluation result 23 based on the corrected selection result 57.

[0114] As described above, in this embodiment, the screen distribution control unit 51 presents the operator OP with the sorting result 57 of true synchronization bursts TNB and false synchronization bursts FNB. Therefore, the operator OP can be notified of the sorting result 57. Furthermore, in this embodiment, the screen distribution control unit 51 accepts an instruction from the operator OP to correct the sorting result 57. Therefore, if the sorting result 57 is incorrect, the operator OP can manually correct the sorting result 57.

[0115] In addition, the threshold strength THI of the detection condition 41, the threshold detection number THND, threshold ratio THP1, and threshold interval THIV of the tentative detection condition 42, as well as the threshold density THD, threshold time THT, and threshold ratio THP2 of the sorting condition 43, especially the threshold density THD, threshold time THT, and threshold ratio THP2 of the sorting condition 43, may be configured so that the operator OP can change the settings.

[0116] A method for detecting spikes from the high-frequency measurement data 20H may involve extracting data from the high-frequency measurement data 20H in a frequency band with a relatively good signal-to-noise ratio, such as the 1.3 kHz band, using wavelet transform or the like, and then detecting spikes from the extracted data using a threshold method. This reduces the risk of falsely detecting noise as a spike. Alternatively, a waveform related to a typical spike may be prepared as a template, and spikes may be detected by template matching. Furthermore, spikes may be detected using a trained model that has trained waveforms related to typical spikes. The waveform related to a spike prepared as training data for the template or trained model may be an actual waveform or a waveform created by a simulator.

[0117] A typical low-frequency measurement data 20L or a PSD graph 65 may be prepared as a template, and a true synchronization burst TNB and a false synchronization burst FNB may be selected by template matching. The typical low-frequency measurement data 20L or the PSD graph 65 may be actual data or data generated by a simulator.

[0118] The neurons 15 and astrocytes 16 do not have to be derived from the illustrated human iPS cells 17, but may be cultured cells of a laboratory animal such as a rat. Furthermore, the human pluripotent stem cells are not limited to the illustrated human iPS cells 17, but may also be human embryonic stem (ES) cells.

[0119] The drug discovery support server 10 may be installed in a pharmaceutical company or a pharmaceutical development contract organization, or may be installed in a data center independent of the pharmaceutical company or the pharmaceutical development contract organization.

[0120] Instead of distributing screen data of the evaluation result display screen 80 including the evaluation result 23 to the operator terminal 11, the evaluation result 23 itself may be distributed to the operator terminal 11. In this case, the operator terminal 11 generates the evaluation result display screen 80 based on the evaluation result 23 under the control of the browser control unit 72.

[0121] The method of presenting the evaluation result 23 to the operator OP is not limited to the example of delivering screen data. The evaluation result 23 may be presented to the operator OP by printing it on a paper medium, or by attaching it to an e-mail and sending it to the operator terminal 11.

[0122] The hardware configuration of the computer constituting the drug discovery support server 10 according to the technology of the present disclosure can be modified in various ways. For example, the drug discovery support server 10 can be configured with multiple computers separated as hardware in order to improve processing power and reliability. For example, the functions of the request reception unit 45, RW control unit 46, and detection unit 47, and the functions of the provisional detection unit 48, selection unit 49, evaluation unit 50, and screen distribution control unit 51 can be distributed and performed by two computers. In this case, the drug discovery support server 10 is configured by two computers. Some or all of the functions of the drug discovery support server 10 may be performed by the operator terminal 11.

[0123] In this way, the hardware configuration of the computer of the drug discovery support server 10 can be changed as appropriate depending on the required performance such as processing power, safety, reliability, etc. Furthermore, not only the hardware but also APs such as the operating program 40 can be duplicated or stored in multiple storages in order to ensure safety and reliability.

[0124] In each of the above embodiments, the hardware structure of the processing unit that executes various processes, such as the request receiving unit 45, the RW control unit 46, the detection unit 47, the provisional detection unit 48, the selection unit 49, the evaluation unit 50, the screen delivery control unit 51, and the browser control unit 72, can be any of the various processors shown below. As described above, the various processors include the CPUs 32A and 32B, which are general-purpose processors that execute software (the operating program 40 and the evaluation AP 70) 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).

[0125] 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.

[0126] 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.

[0127] Furthermore, more specifically, the hardware structure of these various processors can be an electric circuit (circuitry) that combines circuit elements such as semiconductor elements.

[0128] From the above description, the technology described in the following supplementary paragraphs can be understood.

[0129] [Supplementary Item 1] A drug discovery support device comprising a processor, wherein the processor acquires measurement data of extracellular potentials of neurons, the measurement data including local field potentials, and selects true synchronous bursts and false synchronous bursts based on intensity in a frequency band related to the local field potentials. [Supplementary Item 2] The drug discovery support device according to Supplementary Item 1, wherein the processor performs an extraction process to extract data in a frequency band related to the local field potentials from the measurement data. [Supplementary Item 3] The drug discovery support device according to Supplementary Item 2, wherein the processor performs the extraction process using any one of a filter, a wavelet transform, and a short-time Fourier transform. [Supplementary Item 4] The drug discovery support device according to any one of Supplementary Items 1 to 3, wherein the processor provisionally detects synchronous bursts from the measurement data, and selects true synchronous bursts and false synchronous bursts from the provisionally detected synchronous bursts based on the intensity and preset selection conditions. [Supplementary Item 5] The drug discovery support device according to Supplementary Item 4, wherein the selection conditions include a threshold value related to a difference from a reference intensity, a threshold value related to a duration of the change in intensity, and a threshold value related to a magnitude of the change in intensity. [Supplementary Item 6] The drug discovery support device according to Supplementary Item 4 or Supplementary Item 5, wherein, when selecting the true synchronization burst and the false synchronization burst from the provisionally detected synchronization burst, the intensity when no change related to the true synchronization burst occurs is used as a reference. [Supplementary Item 7] The drug discovery support device according to any one of Supplementary Item 1 to Supplementary Item 6, wherein the processor presents the selection results of the true synchronization burst and the false synchronization burst to an operator. [Supplementary Item 8] The drug discovery support device according to Supplementary Item 7, wherein the processor accepts an operator's instruction to correct the selection results.[Supplementary Item 9] The drug discovery support device according to any one of Supplementary Items 1 to 8, wherein the processor uses a trained model generated from training data including waveforms of the true synchronized bursts selected based on the intensity and waveforms of the false synchronized bursts selected based on the intensity, inputs a waveform, the waveform of which is unknown whether it is a true synchronized burst or a false synchronized burst, to the trained model, and causes the trained model to output a selection result indicating whether it is the true synchronized burst or the false synchronized burst. [Supplementary Item 10] The drug discovery support device according to any one of Supplementary Items 1 to 9, wherein the measurement data is measured using a microelectrode array. [Supplementary Item 11] The drug discovery support device according to any one of Supplementary Items 1 to 10, wherein the nerve cells are derived from human pluripotent stem cells. [Supplementary Item 12] The drug discovery support device according to any one of Supplementary Items 1 to 11, wherein the nerve cells include glutamatergic nerve cells. [Supplementary Item 13] The drug discovery support device according to any one of Supplementary Items 1 to 12, wherein the nerve cells are co-cultured with astrocytes. [Supplementary Item 14] The drug discovery support device according to Supplementary Item 13, wherein the astrocytes are derived from human pluripotent stem cells.

[0130] 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-described 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.

[0131] 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.

[0132] 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."

[0133] 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 drug discovery support device comprising a processor, which acquires measurement data of extracellular potentials of neurons, including local field potentials, and selects true synchronous bursts and false synchronous bursts based on the intensity in a frequency band related to the local field potentials.

2. The drug discovery support device according to claim 1, wherein the processor performs extraction processing to extract data in a frequency band related to the local field potential from the measurement data.

3. The drug discovery support device according to claim 2, wherein the processor performs the extraction process using one of a filter, a wavelet transform, and a short-time Fourier transform.

4. The drug discovery support device of claim 1, wherein the processor provisionally detects synchronization bursts from the measurement data, and selects true synchronization bursts and false synchronization bursts from the provisionally detected synchronization bursts based on the intensity and predetermined selection conditions.

5. A drug discovery support device as described in claim 4, wherein the selection conditions include a threshold value related to the difference between the intensity and a standard, a threshold value related to the duration of the change in intensity, and a threshold value related to the magnitude of the change in intensity.

6. A drug discovery support device as described in claim 4, wherein when sorting the true synchronization burst and the false synchronization burst from the provisionally detected synchronization burst, the intensity when no change related to the true synchronization burst occurs is used as a standard.

7. The drug discovery support device according to claim 1, wherein the processor presents the results of sorting the true synchronous burst and the false synchronous burst to an operator.

8. The drug discovery support device according to claim 7, wherein the processor accepts an operator's instruction to correct the selection results.

9. The drug discovery support device described in claim 1, wherein the processor uses a trained model generated from training data including waveforms of the true synchronization bursts selected based on the intensity and waveforms of the false synchronization bursts selected based on the intensity, inputs a waveform that is unknown as to whether it is the true synchronization burst or the false synchronization burst into the trained model, and outputs a selection result indicating whether it is the true synchronization burst or the false synchronization burst from the trained model.

10. The drug discovery support device according to claim 1, wherein the measurement data is measured using a microelectrode array.

11. The drug discovery support device according to claim 1, wherein the nerve cells are derived from human pluripotent stem cells.

12. The drug discovery support device according to claim 1, wherein the neurons include glutamatergic neurons.

13. The drug discovery support device according to claim 1, wherein the nerve cells are co-cultured with astrocytes.

14. The drug discovery support device according to claim 13, wherein the astrocytes are derived from human pluripotent stem cells.

15. A method for operating a drug discovery support device, comprising: acquiring measurement data of extracellular potentials of neurons, the measurement data including local field potentials; and sorting true synchronous bursts and false synchronous bursts based on the intensity of the measurement data in a frequency band related to the local field potentials.

16. An operating program for a drug discovery support device that causes a computer to perform processing including: acquiring measurement data of the extracellular potential of neurons, the measurement data including local field potential; and sorting true synchronous bursts and false synchronous bursts based on the intensity of the measurement data in a frequency band related to the local field potential.

17. A method for evaluating the effect of a test substance on neurons, comprising: acquiring measurement data of the extracellular potential of neurons, the measurement data including local field potential; and distinguishing between true synchronous bursts and false synchronous bursts based on the intensity of the measurement data in a frequency band related to the local field potential.

18. The evaluation method according to claim 17, which evaluates the effect of a test substance on central neurotoxicity.

Images