Ion channel analysis device

Abstract

Provided is an ion channel analysis device capable of obtaining a mathematical parameter for analyzing a state of an ion channel. Specifically, an ion channel analysis device 1 comprises: a detection unit 20 that detects an ion current I flowing through an ion channel 10; and an analysis unit 30 that acquires first data 40 related to the ion current I detected by the detection unit 20. There is a regression model M using, as teacher data, the known first data 40 and a known mathematical parameter 70 corresponding to the known first data 40. The analysis unit 30 outputs the mathematical parameter 70 corresponding to the input first data 40 by inputting the first data 40 related to the ion current I detected by the detection unit 20 to the regression model M.

Description

Ion channel analyzer

[0001] The present disclosure relates to an ion channel analyzer.

[0002] Transmembrane proteins called ion channels exist in the cell membranes of living organisms (see, for example, Patent Document 1). Ion channels act as pathways for ions in the cell membrane. Ions have an electric charge. Ions passively pass through the cell membrane through ion channels. Ion channels open and close, or open and close.

[0003] When an ion channel opens, the ionic current flowing through it increases. When an ion channel closes, the ionic current flowing through it decreases.

[0004] Special Publication No. 2010-513332

[0005] Incidentally, it is conceivable to analyze the state of an ion channel based on data (such as a current waveform or a histogram) obtained from the ion current flowing through the ion channel.

[0006] In order to analyze the state of an ion channel based on data obtained from the ionic current flowing through the ion channel, it is necessary to obtain mathematical parameters corresponding to the data, but obtaining such mathematical parameters has been difficult.

[0007] The present disclosure has been made in view of the above points, and an object thereof is to provide an ion channel analysis device capable of obtaining mathematical parameters for analyzing the state of an ion channel.

[0008] The ion channel analysis device according to the present disclosure includes a detection unit that detects an ion current flowing through an ion channel, and an analysis unit that obtains first data related to the ion current detected by the detection unit. The device has a regression model in which known first data and known mathematical parameters corresponding to the known first data are used as training data. The analysis unit inputs the first data related to the ion current detected by the detection unit into the regression model, and outputs the mathematical parameters corresponding to the input first data.

[0009] According to this configuration, the regression model uses the known first data and the known mathematical parameters corresponding to the known first data as training data, i.e., the regression model learns the correlation between the known first data and the known mathematical parameters.

[0010] The analysis unit inputs first data relating to the ion current detected by the detection unit to a regression model that reflects a correlation between the known first data and known mathematical parameters, thereby enabling the analysis unit to output mathematical parameters corresponding to the input first data.

[0011] The output mathematical parameters can be used to analyze the state of the ion channel.

[0012] As described above, it is possible to provide an ion channel analysis device capable of obtaining mathematical parameters for analyzing the state of an ion channel.

[0013] In one embodiment, the first data includes a current waveform showing the change over time of the ionic current, and the regression model uses the known current waveform and the known mathematical parameters corresponding to the known current waveform as training data, and the analysis unit inputs the current waveform related to the ionic current detected by the detection unit into the regression model, thereby outputting the mathematical parameters corresponding to the input current waveform.

[0014] According to this configuration, by applying a current waveform as the first data, the analysis unit outputs a mathematical parameter corresponding to the input current waveform, thereby obtaining a mathematical parameter effective for analyzing the state of an ion channel.

[0015] In one embodiment, the first data includes a histogram corresponding to a current waveform showing a time change of the ionic current and showing the frequency of each value of the ionic current, the regression model uses the known histogram and the known mathematical parameters corresponding to the known histogram as training data, and the analysis unit inputs the histogram related to the ionic current detected by the detection unit into the regression model, and outputs the mathematical parameters corresponding to the input histogram.

[0016] According to this configuration, by applying a histogram as the first data, the analysis unit outputs mathematical parameters corresponding to the input histogram. As a result, as in the case of a current waveform, mathematical parameters effective for analyzing the state of an ion channel can be obtained. In particular, when a histogram is applied as the first data, the number of mathematical parameters required for analyzing the state of an ion channel can be reduced compared to when a current waveform is applied as the first data.

[0017] In one embodiment, the analysis unit analyzes the state of the ion channel based on the output mathematical parameters.

[0018] According to this configuration, the task of analyzing the state of ion channels can be automated.

[0019] In one embodiment, the mathematical parameters include information regarding the open / closed state of the ion channel, and the analysis unit analyzes the open / closed state of the ion channel based on the output mathematical parameters.

[0020] With this configuration, it is possible to grasp the open / closed state of the ion channel.

[0021] In one embodiment, the mathematical parameters include information regarding an open probability as the open / closed state of the ion channel, and the analysis unit calculates the open probability of the ion channel based on the output mathematical parameters.

[0022] With this configuration, the open probability of the ion channel can be obtained.

[0023] In one embodiment, the ion current I includes a base current value and a peak current value greater than the base current value, and the mathematical parameters include, as information related to the open probability, the base current value, the peak current value, a time constant from the base current value to the peak current value, a time constant from the peak current value to the base current value, a duration at the base current value, and a duration at the peak current value.

[0024] With this configuration, the open probability of the ion channel can be obtained with high accuracy.

[0025] In one embodiment, the ion channel analysis device is applied to said ion channels formed in an artificial cell membrane.

[0026] This configuration is effective in conducting experiments using artificial cell membranes.

[0027] According to the present disclosure, it is possible to provide an ion channel analysis device capable of obtaining mathematical parameters for analyzing the state of an ion channel.

[0028] FIG. 1 shows an ion channel formed in an artificial cell membrane. FIG. 2 shows a current waveform of an ion current. FIG. 3 shows a histogram of the ion current. FIG. 4 shows a regression model M according to the first embodiment. FIG. 5 shows a first analysis target histogram and a reference histogram according to the second embodiment. FIG. 6 shows a flowchart for determining the number of clusters according to the second embodiment.

[0029] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. The following description of the preferred embodiments is merely exemplary in nature and is not intended to limit the present disclosure, its applications, or its uses.

[0030] <First embodiment> (Ion channel analysis device) An ion channel analysis device 1 according to a first embodiment will be described. The ion channel analysis device 1 is applied to an ion channel 10. The ion channel analysis device 1 analyzes the state of the ion channel 10. In this example, the ion channel 10 is formed in an artificial cell membrane 5.

[0031] FIG. 1 shows an ion channel 10 formed in an artificial cell membrane 5. The artificial cell membrane 5 is an artificial replica of a biological cell membrane. When a water droplet 4 is dropped into an organic solvent (oil) in which lipid molecules 2 are dispersed, a monolayer of lipid molecules 2 spontaneously forms around the water droplet 4. When two water droplets 4 are brought into contact with each other, a monolayer of lipid molecules 2 overlaps at the contact point between the two water droplets 4. An artificial cell membrane 5 composed of a bilayer of lipid molecules 2 is formed between the two water droplets 4. One of the two water droplets 4 mimics the interior of a biological cell, and the other of the two water droplets 4 mimics the exterior of a biological cell.

[0032] Proteins 3, which are the basis of ion channels 10, are dispersed in the water droplets 4. When the proteins 3 attach to the artificial cell membrane 5, the ion channels 10 are formed.

[0033] The ion channel 10 is a passage that penetrates the artificial cell membrane 5. The ion channel 10 connects the two water droplets 4. Ions passively pass through the ion channel 10 between the two water droplets 4. The ion channel 10 is a passage for ions in the artificial cell membrane 5. Ions have an electric charge. Examples of ions include potassium ions and sodium ions.

[0034] The ion channel 10 opens and closes, opening and closing. When the ion channel 10 is open, the ionic current I flowing through the ion channel 10 between the two water droplets 4 increases. When the ion channel 10 is closed, the ionic current I flowing through the ion channel 10 between the two water droplets 4 decreases.

[0035] The ion channel analysis device 1 includes a detection unit 20 and an analysis unit 30. The detection unit 20 is a known ammeter. The detection unit 20 detects an ion current I flowing through the ion channel 10. The detection unit 20 is connected in series to both ends of the ion channel 10 (one water droplet 4 side and the other water droplet 4 side). The detection unit 20 also has a function of applying a command voltage to the ion channel 10.

[0036] The analysis unit 30 is connected to the detection unit 20. The analysis unit 30 is built into the main body of the ion channel analysis device 1. The analysis unit 30 is, for example, a computer. The analysis unit 30 includes, for example, a processor mounted on a board and a memory device that stores software for operating the processor. The analysis unit 30 performs the arithmetic processing described below.

[0037] The analysis unit 30 obtains first data 40 relating to the ion current I detected by the detection unit 20. The analysis unit 30 analyzes the state of the ion channel 10 based on the first data 40 relating to the ion current I. In particular, in this example, the analysis unit 30 analyzes the open / closed state of the ion channel 10.

[0038] (Current Waveform) The first data 40 includes a current waveform 50. The current waveform 50 shows the change over time of the ionic current I. FIG. 2 shows the current waveform 50 of the ionic current I. In FIG. 2, the horizontal axis shows time t, and the vertical axis shows the value of the ionic current I. The unit of time is, for example, [s]. The unit of the ionic current I is, for example, [A]. The current waveform 50 is a collection of sets of data for time t and data for the ionic current I.

[0039] In the current waveform 50, the ionic current I includes a leakage current value IL as a base current value and a peak current value IP. The leakage current value IL corresponds to the value of the ionic current I when the ion channel 10 is closed. Even when the ion channel 10 is closed, the ionic current I does not become completely zero, but rather flows slightly through the ion channel 10 by leaking (leaking) through the artificial cell membrane 5. Therefore, the leakage current value IL is slightly greater than zero.

[0040] The magnitude of the leakage current value IL in the ion current I is determined in proportion to the magnitude of the command voltage applied to the ion channel 10. If the applied command voltage is a negative potential, the leakage current value IL in the ion current I also becomes a negative value.

[0041] The peak current value IP is greater in absolute value than the leakage current value IL. The peak current value IP corresponds to the value of the ionic current I when the ion channel 10 is open. When the ion channel 10 is open, a large amount of the ionic current I flows through the ion channel 10. Therefore, the peak current value IP is greater in absolute value than the leakage current value IL. According to FIG. 2, the ionic current I alternates between the leakage current value IL and the peak current value IP. The leakage current value IL is not a completely constant value, but fluctuates randomly within the range of the noise σ. The peak current value IP is not a completely constant value, but fluctuates randomly within the range of the noise σ.

[0042] When the ion channel 10 changes from closed to open, that is, when the ion current I reaches the peak current value IP from the leakage current value IL, it takes a time equal to the first time constant τo. When the ion channel 10 changes from open to closed, that is, when the ion current I reaches the peak current value IP from the leakage current value IL, it takes a time equal to the second time constant τc.

[0043] The ionic current I continues at the leakage current value IL for the closed state duration Tc. The closed state durations Tc at the leakage current values ​​IL of the first to fourth valleys are indicated by Tc1, Tc2, Tc3, and Tc4. The total closed state duration Tc for the target period is the sum of Tc1 to Tc4. In other words, Tc = Tc1 + Tc2 + Tc3 + Tc4.

[0044] The ionic current I continues at the peak current value IP for the open duration To. The open durations To at the peak current values ​​IP of the first to fourth peaks are indicated by To1, To2, To3, and To4. The total open duration To during the target period is the sum of To1 to To4. In other words, To = To1 + To2 + To3 + To4.

[0045] The open state of the ion channel 10 is represented by an open probability Po. The open probability Po is the ratio of the open duration To during which the ion channel 10 is open to the target period. The open probability Po is roughly given by To / (Tc+To) (Po=To / (Tc+To)).

[0046] The ion channel 10 has a closed probability Pc as an open or closed state. The closed probability Pc is the ratio of the closed duration Tc during which the ion channel 10 is closed to the target period. The closed probability Pc is roughly calculated as Tc / (Tc+To) (Pc=Tc / (Tc+To)).

[0047] The sum of the open probability Po and the closed probability Pc is 1 (Po + Pc = 1). If one of the open probability Po and the closed probability Pc is determined, the other is also determined. Calculating the open probability Po and calculating the closed probability Pc are essentially the same. The open probability Po will be explained below.

[0048] The target period includes the closed duration Tc and the open duration To as well as the first time constant τo and the second time constant τc. In order to accurately calculate the open probability Po, the first time constant τo and the second time constant τc must be allocated to either the closed duration Tc or the open duration To.

[0049] For example, an appropriate threshold value IA is set for the ion current I. It is conceivable to allocate the time when the ion current I is smaller than the threshold value IA to the closed duration Tc, and allocate the time when the ion current I is larger than the threshold value IA to the open duration To.

[0050] A human can simply visually set the threshold value IA between the leakage current value IL and the peak current value IP for a given current waveform 50. This allows a human to accurately determine the open probability Po, taking into account the first time constant τo and the second time constant τc.

[0051] However, it is difficult for a computer to automatically calculate an accurate open probability Po based on the current waveform 50. In the first place, a computer cannot recognize the leakage current value IL and the peak current value IP even when the current waveform 50 is given. For this reason, the computer cannot appropriately set the threshold value IA between the leakage current value IL and the peak current value IP for each given current waveform 50. Furthermore, the influence of noise σ also makes it difficult for a computer to recognize the leakage current value IL and the peak current value IP.

[0052] Although it is conceivable to fix the threshold value IA in advance at an appropriate predetermined value, the state of the current waveform 50 will differ each time depending on the type of ion channel 10 and the magnitude of the command voltage applied to the ion channel 10. For example, both the leakage current value IL and the peak current value IP may increase or decrease (the entire current waveform 50 may move up or down). For this reason, if the threshold value IA is fixed in advance at an appropriate predetermined value, both the leakage current value IL and the peak current value IP in the current waveform 50 may increase or decrease relative to the threshold value IA, making it impossible to obtain an accurate open probability Po.

[0053] As such, it is difficult to automatically and accurately determine the open probability Po by a computer based on the current waveform 50.

[0054] (Histogram) The first data 40 includes a histogram 60. The histogram 60 corresponds to the current waveform 50. Specifically, the histogram 60 is converted from the current waveform 50. The histogram 60 indicates a frequency F for each value of the ionic current I. FIG. 3 shows the histogram 60 of the ionic current I. In FIG. 2, the horizontal axis indicates the ionic current I, and the vertical axis indicates the frequency F. The unit of the frequency F is dimensionless [-]. The frequency F is also a probability.

[0055] The histogram 60 is essentially a bar graph, but by increasing the number of data points, it becomes a curved line as shown in Figure 3. In this example, the histogram 60 is configured as a Gaussian distribution, which is also known as a normal distribution.

[0056] In the histogram 60, a closed-side peak 61 corresponding to the closed duration Tc (leakage current IL) is formed on the side (left side) where the ionic current I is small, and an open-side peak 62 corresponding to the open duration To (peak current IP) is formed on the side (right side) where the ionic current I is large. In the histogram 60, a skirt 63 corresponding to the first time constant τo, the second time constant τc, and the noise σ is formed between the closed-side peak 61 and the open-side peak 62.

[0057] The closed side mountain portion 61 has a closed area A1. The open side mountain portion 62 has an open area A2. The skirt portion 63 has a small skirt area A3.

[0058] The open probability Po of the ion channel 10 is roughly given by A2 / (A1+A2) (Po=A2 / (A1+A2)), and the close probability Pc of the ion channel 10 is roughly given by A1 / (A1+A2) (Pc=A1 / (A1+A2)).

[0059] As described above, since obtaining the open probability Po and obtaining the closed probability Pc are essentially the same, the open probability Po will be explained below.

[0060] To accurately calculate the open probability Po, the tail area A3 must be allocated to either the closed area A1 or the open area A2. For example, an appropriate threshold IA is set for the ionic current I. It is conceivable to allocate the area when the ionic current I is smaller than the threshold IA to the closed area A1, and the area when the ionic current I is larger than the threshold IA to the open area A2.

[0061] A human can visually distinguish between the closed peaks 61 and the open peaks 62 for a given histogram 60 and set the threshold value IA midway between them (the tail 63) each time. This allows a human to calculate an accurate open probability Po that also takes into account the tail area A3 of the tail 63.

[0062] However, it is difficult for a computer to automatically calculate an accurate open probability Po based on the histogram 60. Even if the histogram 60 is given, the computer cannot distinguish between the closed peak portion 61 and the open peak portion 62. For this reason, the computer cannot appropriately set the threshold value IA to the midpoint (the skirt portion 63) between the closed peak portion 61 and the open peak portion 62 each time.

[0063] It is conceivable to fix the threshold value IA in advance at an appropriate predetermined value, but the appearance of the histogram 60 will differ depending on the type of ion channel 10 and the magnitude of the command voltage applied to the ion channel 10. For example, when both the leakage current value IL and the peak current value IP increase or decrease, the entire histogram 60 will shift to the right or left. For this reason, when the threshold value IA is fixed in advance at an appropriate predetermined value, both the closed-side peak portion 61 and the open-side peak portion 62 in the histogram 60 may increase or decrease compared to the threshold value IA (they may be located to the right or left), making it impossible to obtain an accurate open probability Po.

[0064] As such, it is difficult to automatically calculate an accurate open probability Po based on the histogram 60 using a computer.

[0065] (Regression Model) Fig. 4 shows the regression model M according to the first embodiment. The analysis unit 30 of the ion channel analysis device 1 uses the regression model M. The regression model M is in the external server 6. Specifically, the regression model M is stored in the external server 6. The analysis unit 30 accesses the regression model M in the external server 6 to input data to the regression model M and receive data output by the regression model M.

[0066] The regression model M is a supervised machine learning model. Specific examples of the regression model M include ridge regression, a gradient boosting decision tree (GDBT), a multi-layer perceptron (MLP), a convolutional neural network (CNN) model, and a transformer model.

[0067] The regression model M is trained as follows: The regression model M uses the known first data 40 and the known mathematical parameters 70 as training data. The known mathematical parameters 70 correspond to the known first data 40.

[0068] When the regression model M is trained, the known first data 40 is provided to the input layer of the regression model M as training data, and the known mathematical parameters 70 are provided to the output layer of the regression model M as training data.

[0069] For example, the regression model M uses a known current waveform 50 and known mathematical parameters 70 as training data. At this time, the known mathematical parameters 70 correspond to the known current waveform 50. Furthermore, the known current waveform 50 is provided as training data to an input layer of the regression model M, and the known mathematical parameters 70 are provided as training data to an output layer of the regression model M.

[0070] Alternatively, the regression model M uses the known histogram 60 and the known mathematical parameters 70 as training data. In this case, the known mathematical parameters 70 correspond to the known histogram 60. Furthermore, the known histogram 60 is provided as training data to an input layer of the regression model M, and the known mathematical parameters 70 are provided as training data to an output layer of the regression model M.

[0071] The mathematical parameters 70 are input variables corresponding to the first data 40 (current waveform 50 and histogram 60). The known first data 40 (current waveform 50 and histogram 60) are generated by numerical simulation based on the known mathematical parameters 70. A large number of mathematical parameters 70 are set while changing the conditions of the mathematical parameters 70. Corresponding first data 40 (current waveform 50 and histogram 60) is generated for each mathematical parameter 70 with different conditions. In this way, a large number of sets of the first data 40 (current waveform 50 and histogram 60) and the mathematical parameters 70 are prepared according to the conditions.

[0072] The mathematical parameters 70 include information regarding the open / closed state of the ion channel 10. The open / closed state of the ion channel 10 refers to a state relating to whether the ion channel 10 is open or closed, and the degree of opening or closing of the ion channel 10. Specifically, the mathematical parameters 70 include information regarding the open probability Po as the open / closed state of the ion channel 10.

[0073] More specifically, the mathematical parameters 70 include, as information regarding the open probability Po of the ion channel 10, the leakage current value IL, the peak current value IP, the first time constant τo from the leakage current value IL to the peak current value IP, the second time constant τc from the peak current value IP to the leakage current value IL, the closed duration Tc at the leakage current value IL, and the open duration To at the peak current value IP.

[0074] Furthermore, the mathematical parameters 70 may include the open probability Po itself as information regarding the open probability Po of the ion channel 10 .

[0075] Regardless of whether a known current waveform 50 or a known histogram 60 is used as the training data to be provided to the input layer, it is preferable to provide these known mathematical parameters 70 (items not in parentheses in Figure 4) to the output layer.

[0076] The mathematical parameters 70 may include, as information regarding the open probability Po of the ion channel 10, the number of opening and closing events D (the number of times the ion channel 10 opens and closes during a target period), the opening and closing timing E (the time at which the ion channel 10 opens and closes), the closed duration Tc for each opening and closing event (e.g., Tc1 to Tc4), and the open duration To for each opening and closing event (e.g., To1 to To4).

[0077] When a known current waveform 50 is used as the training data to be provided to the input layer, it is preferable to provide these known mathematical parameters 70 (items in parentheses in Figure 4) to the output layer. However, when a known histogram 60 is used as the training data to be provided to the input layer, it is not essential to provide these known mathematical parameters 70 (items in parentheses in Figure 4) to the output layer.

[0078] The analysis unit 30 executes the regression model M as follows. The analysis unit 30 inputs first data 40 related to the ionic current I detected by the detection unit 20 into the regression model M, and outputs mathematical parameters 70 corresponding to the input first data 40. The analysis unit 30 accesses the regression model M stored in the external server 6, and inputs the first data 40 related to the detected ionic current I into the regression model M. The analysis unit 30 acquires the mathematical parameters 70 (corresponding to the first data 40) output by the regression model M in the external server 6. The analysis unit 30 outputs the acquired mathematical parameters 70.

[0079] As the mathematical parameter 70 corresponding to the first data 40, information relating to the open probability Po (open / closed state) of the ion channel 10 is output.

[0080] For example, the analysis unit 30 inputs a current waveform 50 relating to the ionic current I detected by the detection unit 20 into the regression model M, and outputs a mathematical parameter 70 corresponding to the input current waveform 50 .

[0081] The leakage current value IL, the peak current value IP, the first time constant τo, the second time constant τc, the close duration Tc, the open duration To, and the open probability Po itself are output as the mathematical parameters 70 corresponding to the current waveform 50. Furthermore, the number of opening and closing events D, the opening and closing timing E, the close duration Tc for each opening and closing event (e.g., Tc1 to Tc4, etc.), and the open duration To for each opening and closing event (e.g., To1 to To4, etc.) are output as the mathematical parameters 70 corresponding to the current waveform 50.

[0082] Alternatively, the analysis unit 30 inputs a histogram 60 relating to the ion current I detected by the detection unit 20 into the regression model M, and outputs mathematical parameters 70 corresponding to the input histogram 60 .

[0083] As mathematical parameters 70 corresponding to the histogram 60, the leakage current value IL, the peak current value IP, the first time constant τo, the second time constant τc, the closed duration Tc, the open duration To, and the open probability Po itself are output.

[0084] The analysis unit 30 analyzes the state of the ion channel 10 based on the output mathematical parameters 70. Specifically, the analysis unit 30 analyzes the open / closed state of the ion channel 10 based on the output mathematical parameters 70.

[0085] More specifically, the analysis unit 30 calculates the open probability Po of the ion channel 10 based on the output mathematical parameters 70. Since specific numerical values ​​are given as the mathematical parameters 70, the analysis unit 30 can easily calculate the open probability Po.

[0086] For example, the analysis unit 30 calculates the open probability Po based on mathematical parameters 70 corresponding to the histogram 60 (see FIG. 3 ). The mathematical parameters 70 corresponding to the histogram 60 give the leakage current value IL and the peak current value IP as specific numerical values. The analysis unit 30 calculates the intermediate value between the leakage current value IL and the peak current value IP, and sets this intermediate value as the threshold value IA.

[0087] The analysis unit 30 allocates the area of ​​the portion of the histogram 60 where the ionic current I is smaller than the threshold value IA (left side) to the closed area A1 of the closed peak 61, and allocates the area of ​​the portion of the histogram 60 where the ionic current I is larger than the threshold value IA (right side) to the open area A2 of the open peak 62. At this time, the base area A3 of the base 63 is allocated to the closed area A1 and the open area A2. This allows the analysis unit 30 to easily calculate the open probability Po by Po = A2 / (A1 + A2).

[0088] Furthermore, for the mathematical parameters 70 corresponding to the histogram 60, the open probability Po itself is given by the regression model M. The analysis unit 30 may apply the open probability Po as the mathematical parameters 70 given by the regression model M as is.

[0089] For example, the validity (reliability) of the mathematical parameter 70 obtained from the regression model M may be confirmed by comparing the open probability Po calculated using the above-mentioned threshold IA with the open probability Po as the mathematical parameter 70 given by the regression model M.

[0090] Alternatively, the analysis unit 30 may calculate the open probability Po based on the mathematical parameters 70 corresponding to the current waveform 50 (see FIG. 2). Note that when the open probability Po is calculated based on the mathematical parameters 70 corresponding to the current waveform 50, the number of items in the mathematical parameters 70 is greater (see FIG. 4) than when the open probability Po is calculated based on the mathematical parameters 70 corresponding to the histogram 60, making the calculation more complex.

[0091] When based on the mathematical parameters 70 corresponding to the current waveform 50, the open probability Po as the mathematical parameters 70 given by the regression model M may be applied as is, just as when based on the mathematical parameters 70 corresponding to the histogram 60.

[0092] (Action and effect of the first embodiment) In the ion channel analysis device 1 according to this embodiment, the detection unit 20 detects the ion current I flowing through the ion channel 10, and the analysis unit 30 obtains the first data 40 detected by the detection unit 20.

[0093] The regression model M uses, as training data, the known first data 40 and the known mathematical parameters 70 corresponding to the known first data 40. In other words, the regression model M is made to learn the correlation between the known first data 40 and the known mathematical parameters 70.

[0094] The analysis unit 30 inputs first data 40 relating to the ion current I detected by the detection unit 20 to a regression model M that reflects the correlation between the known first data 40 and known mathematical parameters 70. This allows the analysis unit 30 to output mathematical parameters 70 that correspond to (correlate with) the input first data 40.

[0095] The state of the ion channel 10 can be analyzed based on the output mathematical parameters 70 .

[0096] As described above, it is possible to provide an ion channel analysis device 1 that can obtain mathematical parameters 70 for analyzing the state of the ion channel 10 .

[0097] The regression model M is a learning model, rather than a data set that simply collects known first data 40 and known mathematical parameters 70. Because the regression model M is a learning model, even if the first data 40 input to the regression model M at the time of execution is complex (for example, even if the noise σ of the current waveform 50 is large as shown in FIG. 2 or even if the tail 63 of the histogram 60 is larger than zero as shown in FIG. 3), the regression model M, whose accuracy has been honed through learning, outputs reliable and appropriate mathematical parameters 70.

[0098] By applying the current waveform 50 as the first data 40, the analysis unit 30 outputs mathematical parameters 70 corresponding to the input current waveform 50. This makes it possible to obtain mathematical parameters 70 that are effective for analyzing the state of the ion channel 10.

[0099] By applying the histogram 60 as the first data 40, the analysis unit 30 outputs mathematical parameters 70 corresponding to the input histogram 60. As a result, similar to the case of the current waveform 50, it is possible to obtain mathematical parameters 70 that are effective for analyzing the state of the ion channel 10.

[0100] In particular, when the histogram 60 is used as the first data 40, the number of items of the mathematical parameters 70 required to analyze the state of the ion channel 10 can be reduced compared to when the current waveform 50 is used as the first data 40 (see FIG. 4 ). This makes it easier to analyze the state of the ion channel 10 based on the mathematical parameters 70.

[0101] The analysis unit 30 itself analyzes the state of the ion channel 10 based on the mathematical parameters 70, thereby automating the task of analyzing the state of the ion channel 10.

[0102] Since the mathematical parameters 70 include information regarding the open / closed state of the ion channel 10, the open / closed state of the ion channel 10 can be grasped.

[0103] In particular, since the mathematical parameters 70 include information regarding the open probability Po of the ion channel 10 as the open or closed state, the open probability Po of the ion channel 10 can be obtained.

[0104] The mathematical parameters 70 include, as information related to the open probability Po, the leakage current value IL, the peak current value IP, the first time constant τo, the second time constant τc, the closed state duration Tc, and the open state duration To. This makes it possible to accurately obtain the open probability Po of the ion channel 10 based on the mathematical parameters 70.

[0105] The ion channel analysis device 1 is applied to the ion channel 10 formed in the artificial cell membrane 5, and is therefore effective in conducting experiments (for medicine, etc.) using the artificial cell membrane 5.

[0106] (Modification of First Embodiment) In the above embodiment, the threshold value IA is used to calculate the open probability Po of the ion channel 10, but the present invention is not limited to this, and the open probability Po may be calculated by other methods.

[0107] In the above embodiment, the open probability Po of the ion channel 10 is obtained based on the mathematical parameters 70. However, the present invention is not limited to this. It is also possible to obtain the close probability Pc based on the mathematical parameters 70 (obtaining the open probability Po and obtaining the close probability Pc are essentially the same thing), to analyze other open / closed states of the ion channel 10 (for example, the degree of opening or closing of the ion channel 10), or to analyze other states of the ion channel 10 (for example, the conductance of the ion channel 10, which can be calculated from the applied voltage and the observed ionic current).

[0108] In the above embodiment, the analysis unit 30 analyzes the state of the ion channel 10 based on the output mathematical parameters 70. However, this is not limiting. The analysis unit 30 may output the mathematical parameters 70, and then a person may analyze the state of the ion channel 10 based on the mathematical parameters 70.

[0109] In the above embodiment, the current waveform 50 and the histogram 60 are exemplified as the first data 40. However, the first data 40 is not limited to these. For example, the first data 40 may be a conductance (unit: siemens) obtained by dividing the ion current I by the command voltage.

[0110] The regression model M may be stored in the memory of the analysis unit 30 instead of being stored in the external server 6 .

[0111] Second Embodiment (Gaussian Mixture Model) An ion channel analysis device 1 according to a second embodiment will be described. In the following description, the same components as those in the above embodiment will be denoted by the same reference numerals, and detailed description thereof may be omitted. Figure 5 shows a first analysis target histogram 80 and a reference histogram 90 according to the second embodiment.

[0112] The detection unit 20 detects an ionic current I flowing through the ion channel 10. The analysis unit 30 obtains a first analyzed histogram 80 based on a current waveform 50 relating to the ionic current I detected by the detection unit 20. The current waveform 50 indicates the change in the ionic current I over time (see FIG. 2 ).

[0113] The first analysis target histogram 80 is the same as the histogram 60 according to the first embodiment (see FIG. 3 ). However, as will be described in detail later, the first analysis target histogram 80 has a smaller number of data points than the second analysis target histogram 85, which will be described later.

[0114] The first analyzed histogram 80 corresponds to the current waveform 50. Specifically, the first analyzed histogram 80 is converted from the current waveform 50. The first analyzed histogram 80 shows the frequency F for each value of the ionic current I.

[0115] The first analyzed histogram 80 is configured with a Gaussian distribution (normal distribution). The first analyzed histogram 80 has a closed-side peak 81 corresponding to the closed duration Tc (leakage current IL) on the side (left) where the ionic current I is small, and an open-side peak 82 corresponding to the open duration To (peak current IP) on the side (right) where the ionic current I is large. In the first analyzed histogram 80, a tail 83 corresponding to the first time constant τo, the second time constant τc, and the noise σ is formed between the closed-side peak 81 and the open-side peak 82.

[0116] When a histogram has a Gaussian distribution, such as the first histogram 80, the number of peaks, such as the closed peak 81 and the open peak 82, in the first histogram 80 is called the number of clusters N. The number of clusters N in the first histogram 80 is 2 (N=2).

[0117] In order to analyze the state of the ion channel 10 (particularly, the open / closed state), the number of clusters N of the histogram must be determined. For example, when the number of clusters N is 2, it can be seen that there are clusters corresponding to when the ion channel 10 is closed (e.g., the closed-side hill 81) and clusters corresponding to when the ion channel 10 is open (e.g., the open-side hill 82). If the number of clusters N is known, the state of the ion channel 10 (particularly, the open / closed state) can be known.

[0118] As can be seen from Figure 5, a human can visually inspect the first histogram 80 and distinguish between the closed peaks 81 and the open peaks 82, and can therefore easily determine that the number of clusters N in the first histogram 80 is 2.

[0119] However, even when a computer is given the first histogram 80 to be analyzed, it cannot distinguish between the closed peaks 81 and the open peaks 82. For this reason, it is difficult for a computer to determine the number of clusters N in the first histogram 80 to be analyzed, and some kind of ingenuity is required.

[0120] The analysis unit 30 of the ion channel analysis device 1 uses a Gaussian mixture model G. The Gaussian mixture model G is stored in the external server 6. Specifically, the Gaussian mixture model G is stored in the external server 6. The analysis unit 30 accesses the Gaussian mixture model G in the external server 6 to input data to the Gaussian mixture model G and to receive data output by the Gaussian mixture model G.

[0121] The Gaussian mixture model G is an unsupervised machine learning model. The Gaussian mixture model G is one of clustering methods, and is a method for analyzing data by assuming that the data is generated by multiple Gaussian distributions (normal distributions).

[0122] The Gaussian mixture model G in the external server 6 has a plurality of reference histograms 90. The plurality of reference histograms 90 are generated by the Gaussian mixture model G. The reference histograms 90 are configured with Gaussian distributions for each number N of clusters by the Gaussian mixture model G. In other words, the reference histograms 90 are generated for each number N of clusters by the Gaussian mixture model G.

[0123] In this example, the Gaussian mixture model G generates five reference histograms 90 configured with Gaussian distributions with the number of clusters N ranging from 1 to 5. If only one ion channel 10 exists, the number of clusters N of 2 is sufficient, but if two or more ion channels 10 may exist, the number of clusters N must be three or more.

[0124] The analysis unit 30 inputs the first histogram to be analyzed 80 to a Gaussian mixture model G in the external server 6. The Gaussian mixture model G generates and outputs a plurality of reference histograms 90 corresponding to the input first histogram to be analyzed 80. The analysis unit 30 acquires the plurality of reference histograms 90 output by the Gaussian mixture model G.

[0125] The analysis unit 30 compares the first histogram to be analyzed 80 with the reference histogram 90 for each number of clusters N. Specifically, the analysis unit 30 compares the first histogram to be analyzed 80 with the reference histogram 90 for which the number of clusters N is 1 (N=1), the reference histogram 90 for which the number of clusters N is 2 (N=2), the reference histogram 90 for which the number of clusters N is 3 (N=3), the reference histogram 90 for which the number of clusters N is 4 (N=4), and the reference histogram 90 for which the number of clusters N is 5 (N=5), in that order.

[0126] The analysis unit 30 determines the number of clusters N when the first histogram to be analyzed 80 is closest to the reference histogram 90 as the number of clusters N of the first histogram to be analyzed 80. For example, the analysis unit 30 calculates the degree of agreement between the first histogram to be analyzed 80 and the reference histogram 90 for each number of clusters N, and determines the number of clusters N when the degree of agreement is highest (highest degree of agreement, smallest difference, etc.).

[0127] In this example, the first histogram to be analyzed 80 is closest to the reference histogram 90, which has a cluster number N of 2 (N=2). The analysis unit 30 determines that the cluster number N of the first histogram to be analyzed 80 is 2 (N=2).

[0128] Furthermore, the analysis unit 30 determines whether the number of clusters N in the first histogram to be analyzed 80 is 2 or more (N≧2?).

[0129] The analysis unit 30 determines that the ion channel 10 is open or closed when the number of clusters N in the first analyzed histogram 80 is 2 or more (N≧2). This is because when the number of clusters is 2 or more, it is clear that there are at least a cluster (e.g., a closed-side hill 81) corresponding to when the ion channel 10 is closed and a cluster (e.g., an open-side hill 82) corresponding to when the ion channel 10 is open. Note that "the ion channel 10 is open or closed" means that the ion channel 10 has two states, an open state and a closed state.

[0130] When the number of clusters N in the first histogram 80 to be analyzed is less than 2 (N<2), that is, when the number of clusters N in the first histogram 80 to be analyzed is 1 (N=1), the analysis unit 30 determines that the ion channel 10 is not open or closed (the ion channel 10 remains open or closed).

[0131] After determining the number of clusters N (=2) of the first analyzed object histogram 80, the analysis unit 30 obtains a second analyzed object histogram 85 based on the current waveform 50. Specifically, the analysis unit 30 converts the current waveform 50 into the second analyzed object histogram 85. The second analyzed object histogram 85 is the same as the histogram 60 according to the first embodiment (see FIG. 3 ).

[0132] The second analyzed object histogram 85 corresponds to the current waveform 50. The second analyzed object histogram 85 corresponds to the first analyzed object histogram 80. The second analyzed object histogram 85 has a larger number of data points than the first analyzed object histogram 80. In other words, the first analyzed object histogram 80 has a smaller number of data points than the second analyzed object histogram 85.

[0133] Based on the second histogram 85 to be analyzed and the determined number of clusters N (=2), the analysis unit 30 outputs Gaussian distribution mathematical parameters 100 corresponding to the second histogram 85 to be analyzed. The Gaussian distribution mathematical parameters 100 are information about the Gaussian distribution that constitutes the second histogram 85 to be analyzed, such as the mean, variance, standard deviation, and specific gravity.

[0134] The analysis unit 30 may analyze the state of the ion channel 10 (for example, the open probability Po) based on the output Gaussian distribution mathematical parameters 100.

[0135] 6 is a flowchart of the second embodiment for determining the number of clusters N. Starting from the start, in a first step S1, the detection unit 20 detects the ion current I flowing through the ion channel 10.

[0136] In a second step S2, the analysis unit 30 obtains a first analyzed histogram 80 based on a current waveform 50 relating to the ionic current I detected by the detection unit 20. The first analyzed histogram 80 has a smaller number of data points than the second analyzed histogram 85.

[0137] In a third step S3 , the analysis unit 30 inputs the first histogram to be analyzed 80 into a Gaussian mixture model G in the external server 6 .

[0138] In a fourth step S4, the Gaussian mixture model G outputs a plurality of reference histograms 90 corresponding to the first histogram to be analyzed 80 for each number N of clusters.

[0139] In a fifth step S5, the analysis unit 30 acquires a plurality of reference histograms 90 for each number N of clusters output by the Gaussian mixture model G. Then, the analysis unit 30 compares the first histogram to be analyzed 80 with the reference histograms 90 for each number N of clusters.

[0140] In a sixth step S6, the analysis unit 30 determines the number of clusters N (= 1 to 5) when the first analyzed histogram 80 is closest to the reference histogram 90 as the number of clusters N of the first analyzed histogram 80. In this example, the number of clusters N of the first analyzed histogram 80 is determined to be 2 (N = 2).

[0141] In seventh step S7, the analysis unit 30 determines whether the number of clusters N of the first analysis target histogram 80 is 2 or greater (N≧2?). If it is determined that the number of clusters N of the first analysis target histogram 80 is 2 or greater (N≧2), the process proceeds to eighth step S8. If it is determined that the number of clusters N of the first analysis target histogram 80 is less than 2 (N<2), that is, if it is determined that the number of clusters N of the first analysis target histogram 80 is 1 (N=1), the process proceeds to ninth step S9.

[0142] In the eighth step S8 (where the number of clusters N is 2 or more (N≧2)), the analysis unit 30 determines that the ion channel 10 is open or closed. After the eighth step S8 is completed, the process proceeds to the tenth step S10.

[0143] In the ninth step S9 (where the number of clusters N is less than 2 (N<2)), the analysis unit 30 determines that the ion channel 10 is not open or closed (the ion channel 10 is constantly open or constantly closed). After the ninth step S9 is completed, the process proceeds to the tenth step S10.

[0144] In a tenth step S10 , the analysis unit 30 obtains a second analyzed histogram 85 that has a larger number of data points than the first analyzed histogram 80 based on the current waveform 50 .

[0145] In an eleventh step S11, the analysis unit 30 outputs the Gaussian distribution mathematical parameters 100 corresponding to the second analysis target histogram 85 based on the second analysis target histogram 85 and the determined number of clusters N (=2). Then, the process ends.

[0146] (Action and effect of the second embodiment) According to the ion channel analysis device 1 of this embodiment, the number of clusters N of the first histogram to be analyzed 80 can be determined by comparing the first histogram to be analyzed 80 with a reference histogram 90 for each number of clusters N and determining the number of clusters N when the first histogram to be analyzed 80 is closest to the reference histogram 90.

[0147] As described above, the number of clusters N is an effective factor for analyzing the state of the ion channel 10 (particularly the open / closed state).

[0148] As described above, it is possible to provide an ion channel analysis device 1 that can obtain the number of clusters N for analyzing the state of the ion channel 10 .

[0149] The number of clusters N of the first histogram to be analyzed 80 can be grasped without relying on human visual inspection.

[0150] Whether the ion channel 10 is open or closed can be easily determined based on whether the number of clusters N in the first analyzed histogram 80 is 2 or more.

[0151] The Gaussian mixture model G can generate an accurate reference histogram 90.

[0152] By using the first histogram 80 to be analyzed, which has a small amount of data, it is possible to speed up the calculation for determining the number of clusters N. On the other hand, by using the second histogram 85 to be analyzed, which has a large amount of data, it is possible to obtain the Gaussian distribution mathematical parameters 100 with high accuracy.

[0153] The ion channel analysis device 1 is applied to the ion channel 10 formed in the artificial cell membrane 5, and is therefore effective in carrying out experiments using the artificial cell membrane 5.

[0154] (Variant of the second embodiment) As described above, when the number of clusters N in the first analyzed histogram 80 is less than 2 (N<2), that is, when the number of clusters N in the first analyzed histogram 80 is 1 (N=1), the analysis unit 30 determines that the ion channel 10 is not open or closed, that is, the ion channel 10 is constantly open or constantly closed.

[0155] In order to determine whether the ion channel 10 is permanently open or permanently closed, a threshold value IA related to the ion current I may be set for the first analyzed histogram 80, and whether the ion channel 10 is permanently open or permanently closed may be determined depending on whether the ion current I is greater than or smaller than the threshold value IA. In this case, if the ion current I is greater than the threshold value IA, the ion channel 10 is determined to be permanently open, and if the ion current I is smaller than the threshold value IA, the ion channel 10 is determined to be permanently closed.

[0156] In the above embodiment, the analysis unit 30 determines whether the ion channel 10 is open or closed based on whether the number of clusters N in the first analyzed histogram 80 is 2 or greater, but this is not limiting. The analysis unit 30 may only perform the analysis up to determining the number of clusters N. A person may analyze the state of the ion channel 10 based on the number of clusters N determined by the analysis unit 30.

[0157] The Gaussian mixture model G may be stored in the memory of the analysis unit 30 instead of being stored in the external server 6 .

[0158] <Other Embodiments> Although the present disclosure has been described above with reference to preferred embodiments, these descriptions are not limiting and, of course, various modifications, substitutions, and combinations are possible.

[0159] The analysis unit 30 may not be built into the main body of the ion channel analysis device 1, but may be provided outside the main body of the ion channel analysis device 1.

[0160] In the above embodiment, the ion channel analyzer 1 is applied to the artificial cell membrane 5, but is not limited to this and may be applied to, for example, a biological cell membrane or the like.

[0161] In the above embodiment, the case where there is one ion channel 10 has been exemplified, but this is not limiting. There may be multiple ion channels 10. In this case, the current waveform 50 may be a superposition of the currents I flowing through the multiple ion channels 10.

[0162] The present disclosure can be applied to the ion channel analysis device 1, and is therefore extremely useful and has high industrial applicability.

[0163] I: Ion current IL: Leak current value (base current value) IP: Peak current value IA: Threshold t: Time τo: First time constant (time constant) τc: Second time constant (time constant) To: Open duration (duration) Tc: Closed duration (duration) σ: Noise Po: Open probability Pc: Closed probability F: Frequency A1: Closed area A2: Open area A3: Tail area D: Number of opening and closing events E: Opening and closing timing M: Regression model G: Gaussian mixture model N: Number of clusters 1: Ion channel analysis device 2: Lipid molecule 3: Organic solvent 4: Water droplet 5: Artificial cell membrane 6: External server 10: Ion channel 20: Detection unit 30: Analysis unit 40: First data 50: Current waveform 60: Histogram 61: Closed peak 62: Open peak 63: Tail 70: Mathematical parameter 80 First histogram to be analyzed 81: Closed peak 82: Open peak 83: Tail 85: Second histogram to be analyzed 90: Reference histogram 100: Gaussian distribution mathematical parameters (mathematical parameters)

Claims

1. An ion channel analysis device comprising: a detection unit that detects an ion current flowing through an ion channel; and an analysis unit that obtains first data related to the ion current detected by the detection unit, wherein there is a regression model using the known first data and known mathematical parameters corresponding to the known first data as teacher data, and the analysis unit inputs the first data related to the ion current detected by the detection unit into the regression model, thereby outputting the mathematical parameters corresponding to the input first data.

2. The ion channel analysis device of claim 1, wherein the first data includes a current waveform showing a time change of the ionic current, the regression model uses the known current waveform and the known mathematical parameters corresponding to the known current waveform as teacher data, and the analysis unit inputs the current waveform related to the ionic current detected by the detection unit into the regression model, thereby outputting the mathematical parameters corresponding to the input current waveform.

3. The ion channel analysis device of claim 1 or 2, wherein the first data includes a histogram corresponding to a current waveform showing a time change of the ion current and showing a frequency of each value of the ion current, the regression model uses the known histogram and the known mathematical parameters corresponding to the known histogram as teacher data, and the analysis unit inputs the histogram related to the ion current detected by the detection unit into the regression model, and outputs the mathematical parameters corresponding to the input histogram.

4. The ion channel analysis device according to claim 1 or 2, wherein the analysis unit analyzes the state of the ion channel based on the output mathematical parameters.

5. An ion channel analysis device as described in claim 1 or 2, wherein the mathematical parameters include information regarding the open / closed state of the ion channel, and the analysis unit analyzes the open / closed state of the ion channel based on the output mathematical parameters.

6. The ion channel analysis device according to claim 5, wherein the mathematical parameters include information regarding the open probability as the open / closed state of the ion channel, and the analysis unit calculates the open probability of the ion channel based on the output mathematical parameters.

7. The ion channel analysis device of claim 6, wherein the ion current I includes a base current value and a peak current value greater than the base current value, and the mathematical parameters include, as information relating to the open probability, the base current value, the peak current value, a time constant from the base current value to the peak current value, a time constant from the peak current value to the base current value, a duration at the base current value, and a duration at the peak current value, 8. The ion channel analysis device according to claim 1 or 2, which is applied to the ion channel formed in an artificial cell membrane.

Images