JP7405373B2 - Chemical substance screening method, program, control device, and culture observation device - Google Patents

Description

The present invention relates to a chemical substance screening method, program, control device, and culture observation device.
This application claims priority based on Japanese Patent Application No. 2018-189810 filed in Japan on October 5, 2018, the contents of which are incorporated herein.

Traditionally, simple systems have been used for high-throughput screening of chemical substances using cells. For example, many screening methods for low-molecular-weight compounds that exhibit anticancer activity use methods such as MTT (methyl thiazolyl tetrazorium) assay and alamarBlue assay, which quantify the number of surviving cells in one well and calculate the survival rate. It was getting worse.

In recent years, screening methods for drug discovery targets and screening methods for low-molecular-weight compounds have become more complex with the development of equipment and imaging technology. As microscopy technology has come to be applied to screening, analysis of multiple parameters from a single type of cell has become possible. Such a screening method is called high content screening (hereinafter referred to as HCS) (see Non-Patent Document 1).
However, conventional chemical substance screening methods are costly and time consuming.

Krausz, Mol. BioSyst. , Volume 3, Pages 232-240, 2007.

In order to solve the above problem, one embodiment of the present invention acquires a plurality of feature amounts related to the morphology of cells included in microscopic images of cells in a plurality of cell groups to which different chemical substances are added at different concentrations. and a feature point, which is a point specified by a set of values indicating each of the plurality of feature quantities in a feature space, which is a space in which coordinate axes are determined for each type of feature quantity related to the shape of the cell. The distance between the different chemical substances in the morphological change region that is the range of chemical substance concentrations in which the magnitude of change in the feature point with respect to the change in the concentration of the chemical substance added to the cell is greater than or equal to a predetermined value. The method of screening chemical substances comprises determining the similarity between the different chemical substances by comparison .

In order to solve the above problem, one aspect of the present invention is a program that causes a computer to execute steps based on the above chemical substance screening method.

In order to solve the above problem, one embodiment of the present invention calculates a plurality of feature values related to the morphology of cells included in microscopic images of cells in a plurality of cell groups to which different chemical substances are added at different concentrations. a feature quantity calculation unit; and a feature quantity, which is a point specified by a set of values indicating each of the plurality of feature quantities in a feature quantity space, which is a space in which coordinate axes are determined for each type of the plurality of feature quantities related to the morphology of the cell; The distance between the points is defined as the distance between the different chemical substances in the morphological change region, which is the range of the concentration of the chemical substance in which the magnitude of the change in the feature point with respect to the change in the concentration of the chemical substance added to the cell is greater than or equal to a predetermined value. The control device includes a similarity determination unit that determines the similarity between the different chemical substances by comparing the substances .

In order to solve the above problem, one aspect of the present invention is a culture observation apparatus including the above control device.

FIG. 1 is a block diagram showing an overview of a culture observation device used in the present invention. FIG. 2 is a front view of a culture observation device used in the present invention. FIG. 2 is a plan view of a culture observation device used in the present invention. 3 is a flowchart showing an example of an observation operation in an incubator used in the present invention. FIG. 2 is a schematic diagram showing an example of a microscope image captured by the imaging device according to the first embodiment of the present invention. It is a flowchart which shows an example of the operation|movement of screening of a chemical substance by the control apparatus of 1st Embodiment of this invention. FIG. 3 is a diagram showing an example of feature points processed by principal component analysis according to the first embodiment of the present invention. FIG. 3 is a diagram showing an example of a density trajectory according to the first embodiment of the present invention. FIG. 3 is a diagram showing an example of the relationship between the distance from the center of the control region of the plot and the chemical substance concentration according to the first embodiment of the present invention. FIG. 3 is a diagram showing an example of the starting concentration of the shape change region for each chemical substance according to the first embodiment of the present invention. FIG. 3 is a diagram showing an example of the relationship between the distance from the center of the dead cell region and the chemical substance concentration in the plot according to the first embodiment of the present invention. FIG. 3 is a diagram showing an example of morphological change regions for each chemical substance according to the first embodiment of the present invention. 2 is a flowchart illustrating an example of an operation of comparing feature points in a feature space between chemical substances by the control device according to the first embodiment of the present invention. FIG. 3 is a diagram illustrating an example of a method for comparing feature points in a feature space between chemical substances by the control device according to the first embodiment of the present invention. It is a block diagram showing the outline of culture observation device 11a used in a 2nd embodiment of the present invention. It is a flowchart which shows an example of the operation|movement of the analysis of a known chemical substance, and the analysis of a new chemical substance by the control apparatus used for 2nd Embodiment of this invention. FIG. 7 is a diagram showing an example of a concentration trajectory on a high concentration side and a time trajectory at a 50% inhibitory concentration according to a second embodiment of the present invention. FIG. 7 is a diagram showing an example of a concentration trajectory on the low concentration side and a time trajectory at a 50% inhibitory concentration according to the second embodiment of the present invention. It is a figure which shows an example of the time trajectory at 50% inhibitory concentration of the 1st chemical substance based on 2nd Embodiment of this invention. It is a figure which shows an example of the time change of the sum of the line segment on the low concentration side and the sum of the line segment on the high concentration side of the 1st chemical substance based on 2nd Embodiment of this invention. It is a figure which shows an example of the time trajectory at 50% inhibitory concentration of the 2nd chemical substance based on 2nd Embodiment of this invention. It is a figure which shows an example of the time change of the sum of line segments on the low concentration side and the sum of line segments on the high concentration side of the second chemical substance according to the second embodiment of the present invention. It is a flowchart which shows an example of the operation|movement of the determination of a new chemical substance by the control apparatus based on 2nd Embodiment of this invention. It is a figure showing an example of comparison of a new chemical substance and a 1st chemical substance concerning a 2nd embodiment of the present invention. It is a figure showing an example of comparison of a new chemical substance and a 2nd chemical substance concerning a 2nd embodiment of the present invention.

(First embodiment)
≪Culture condition evaluation device≫
The chemical substance screening method of the present invention includes a first step of adding a chemical substance to each cell included in the cell group, and a second step of observing morphological changes in the cells to which the chemical substance has been added. 2 steps are included. Here, the cell group means a plurality of cells. In the first step, different concentrations of chemicals are added to the cells. In the second step, differences in cell morphology changes due to differences in concentration are observed. This second step is performed using a culture condition evaluation device. This culture condition evaluation device is included in a culture observation device. The culture state evaluation device is the control device 41 of the culture observation device 11.
This culture observation device will be explained below.

FIG. 1 is a block diagram showing an outline of a culture observation apparatus used in the present invention. 2 and 3 are a front view and a plan view of the culture observation device used in the present invention.

The culture observation device 11 used in the present invention includes an upper casing 12, a lower casing 13, and a display section 50. In the assembled state of the culture observation device 11, the upper casing 12 is placed on the lower casing 13. Note that the internal spaces of the upper casing 12 and the lower casing 13 are partitioned into upper and lower parts by a base plate 14.

First, an outline of the configuration of the upper casing 12 will be explained. A thermostatic chamber 15 for culturing cells is formed inside the upper casing 12. This constant temperature chamber 15 has a temperature adjustment device 15a and a humidity adjustment device 15b, and the inside of the constant temperature chamber 15 is maintained in an environment suitable for culturing cells (for example, an atmosphere with a temperature of 37° C. and a humidity of 90%). Note that illustrations of the temperature adjustment device 15a and the humidity adjustment device 15b in FIGS. 2 and 3 are omitted).

A large door 16, a middle door 17, and a small door 18 are arranged at the front of the constant temperature room 15. The large door 16 covers the front surfaces of the upper casing 12 and the lower casing 13. The middle door 17 covers the front surface of the upper casing 12 and isolates the constant temperature room 15 from the outside environment when the large door 16 is opened. The small door 18 is a door for carrying in and out a culture container 19 for culturing cells, and is attached to the middle door 17. By carrying in and out the culture container 19 through this small door 18, it becomes possible to suppress environmental changes in the thermostatic chamber 15. The airtightness of the large door 16, middle door 17, and small door 18 is maintained by packings P1, P2, and P3, respectively.

Further, in the constant temperature room 15, a stocker 21, an observation unit 22, a container transport device 23, and a transport table 24 are arranged. Here, the transport table 24 is arranged in front of the small door 18, and the culture container 19 is carried in and out through the small door 18.

Stocker 21 is arranged on the left side of thermostatic chamber 15 when viewed from the front of upper casing 12 (lower side in FIG. 3). The stocker 21 has a plurality of shelves, and each shelf of the stocker 21 can accommodate a plurality of culture containers 19. Note that each culture container 19 accommodates cells to be cultured together with a medium.

The observation unit 22 is arranged on the right side of the thermostatic chamber 15 when viewed from the front of the upper casing 12. This observation unit 22 can perform time-lapse observation of cells within the culture container 19.

Here, the observation unit 22 is fitted into the opening of the base plate 14 of the upper casing 12 and arranged. The observation unit 22 includes a sample stage 31, a stand arm 32 projecting above the sample stage 31, and a main body portion 33 incorporating a microscopic optical system 223 for phase contrast observation and an imaging device 224. The sample stage 31 and stand arm 32 are placed in the thermostatic chamber 15, while the main body portion 33 is housed within the lower casing 13.

The sample stage 31 is made of a translucent material, and the culture container 19 can be placed thereon. This sample stage 31 is configured to be movable in the horizontal direction, and the position of the culture container 19 placed on the top surface can be adjusted. Further, the stand arm 32 has a built-in LED light source 221. The stand arm 32 is provided with an illumination ring diaphragm, and the light intensity distribution of the illumination light from the LED light source 221 irradiated onto the sample stage 31 can be variably adjusted. The stand arm 32 functions as a lighting device 222. The illumination device 222 and the microscopic optical system 223 constitute a phase contrast microscope. The microscopic optical system 223 is provided with a condenser lens, an objective lens, a phase plate, and the like, and is configured similarly to the microscopic optical system of a known phase contrast microscope.

That is, the illumination light emitted by the LED light source 221 is narrowed down by the illumination ring diaphragm provided on the stand arm 32 (illumination device 222). The illumination light focused by the illumination ring diaphragm passes through a condenser lens and reaches the cells in the culture container 19. The illumination light that reaches the cells in the culture container 19 is separated into direct light and diffracted light. Here, the direct light is illumination light that travels straight through the inside of the cells in the culture container 19. Diffracted light is illumination light diffracted by cells, which are phase objects. Since the diffraction phenomenon occurs in areas with different refractive indexes, the diffracted light is transmitted to the shape of the cell, which is a phase object, such as the boundary between the cell and the solution in which the cell is immersed in the culture container 19, and the internal structure of the cell. Contains information.

The direct light and the diffracted light pass through the objective lens and reach the phase plate. The phase plate has a quarter wavelength plate and an ND filter formed in a ring shape, and the portion other than the quarter wavelength plate and the ND filter is transparent. Here, the quarter-wave plate shifts the phase of light by a quarter wavelength. ND filters absorb light.
The direct light is focused by the objective lens and passes through the ring-shaped quarter wavelength plate of the phase plate, so the phase is shifted by a quarter wavelength. Further, the brightness of the direct light passing through the quarter-wave plate is weakened. On the other hand, most of the diffracted light passes through the transparent portion of the phase plate, and its phase and brightness do not change.
The direct light and the diffracted light reach the image plane of the microscopic optical system 223 and form an image with a contrast of brightness and darkness.

The imaging device 224 can obtain a microscopic image of the cells by imaging the cells in the culture container 19 that are transmitted and illuminated from above the sample stage 31 by the stand arm 32 through the microscopic optical system 223 .

The container conveyance device 23 is arranged at the center of the thermostatic chamber 15 when viewed from the front of the upper casing 12. This container transport device 23 transfers the culture container 19 between the stocker 21 , the sample stage 31 of the observation unit 22 , and the transport stage 24 .

As shown in FIG. 3, the container transport device 23 includes a vertical robot 34 having a multi-joint arm, a rotation stage 35, a mini stage 36, and an arm portion 37. The rotation stage 35 is attached to the tip of the vertical robot 34 via a rotation shaft 35a so as to be rotatable by 180 degrees in the horizontal direction. Therefore, in the rotation stage 35, the arm portions 37 can be opposed to the stocker 21, the sample stage 31, and the transport stage 24, respectively.

Further, the mini stage 36 is attached to the rotation stage 35 so as to be slidable in the horizontal direction. An arm portion 37 that grips the culture container 19 is attached to the mini stage 36. The arm section 37 carries out the gripped culture container 19 from the stocker 21 and places it on the sample stage 31 of the observation unit 22.

Next, an outline of the configuration of the lower casing 13 will be explained. Inside the lower casing 13, a main body portion 33 of the observation unit 22 and a control device 41 of the culture observation apparatus 11 are housed.

The control device 41 is connected to the temperature adjustment device 15a, the humidity adjustment device 15b, the observation unit 22, and the container conveyance device 23, respectively. This control device 41 centrally controls each part of the culture observation device 11 according to a predetermined program.

As an example, the control device 41 controls the temperature adjustment device 15a and the humidity adjustment device 15b to maintain the inside of the constant temperature room 15 at predetermined environmental conditions. Furthermore, the control device 41 controls the observation unit 22 and the container transport device 23 based on a predetermined observation schedule to automatically execute the observation sequence for the culture container 19. Further, the control device 41 (culture state evaluation device) executes a culture state evaluation process for evaluating the culture state of cells based on the images acquired in the observation sequence.

Here, the control device 41 (cultivation state evaluation device) includes a CPU 42 and a storage section 43. The CPU 42 is a processor that executes various calculation processes of the control device 41 (cultivation state evaluation device). The CPU 42 is configured of an integrated circuit such as an LSI (Large Scale Integration), for example. Note that part or all of the CPU 42 may be configured as software. That is, the CPU 42 may be configured by combining hardware and software. Further, the CPU 42 functions as a feature value calculation section 44, a coordinate calculation section 45, a density trajectory calculation section 46, a change area determination section 47, and a similarity determination section 48, respectively, by executing the program. Note that the operations of the feature quantity calculation section 44, coordinate calculation section 45, concentration locus calculation section 46, change area determination section 47, and similarity determination section 48 will be described later.

The storage unit 43 includes a hard disk, a nonvolatile storage medium such as a flash memory, and the like. This storage unit 43 stores management data regarding each culture container 19 stored in the stocker 21, data on microscopic images captured by the imaging device, and data on known chemical substances. Furthermore, the storage unit 43 stores programs executed by the CPU 42.

The above management data includes (a) index data indicating each culture container 19, (b) storage position of the culture container 19 in the stocker 21, (c) type and shape of the culture container 19 (well plate, (d) types of cells cultured in the culture container 19, (e) observation schedule for the culture container 19, (f) imaging conditions during time-lapse observation (magnification of objective lens, observation inside the container) location, etc.), etc. Furthermore, regarding the culture container 19 such as a well plate in which cells can be simultaneously cultured in a plurality of small containers, management data is generated for each small container.

The display unit 50 includes a liquid crystal display and the like, and displays images output by the control device 41 (cultivation state evaluation device). The image displayed by the display unit 50 includes an image of the trajectory calculated by the concentration trajectory calculation unit 46 of the control device 41 (culture state evaluation device). Furthermore, the image displayed by the display unit 50 includes the result determined by the similarity determination unit 48 of the control device 41 (cultivation state evaluation device).

≪Example of observation operation in culture condition evaluation device≫
Next, an example of the observation operation in the culture observation apparatus 11 used in the present invention will be described with reference to the flowchart in FIG. 4. FIG. 4 shows an example of an operation in which the culture container 19 carried into the thermostatic chamber 15 is subjected to time-lapse observation according to a registered observation schedule.

Step S101: The CPU 42 compares the observation schedule of the management data in the storage unit 43 with the current date and time, and determines whether the observation start time for the culture container 19 has arrived. When the observation start time has come (step S101; YES), the CPU 42 shifts the process to step S102. On the other hand, if it is not the observation time for the culture container 19 (step S101; NO), the CPU 42 waits until the time of the next observation schedule.

Step S102: The CPU 42 instructs the container transport device 23 to transport the culture container 19 corresponding to the observation schedule. Then, the container transport device 23 carries out the instructed culture container 19 from the stocker 21 and places it on the sample stage 31 of the observation unit 22. Note that when the culture container 19 is placed on the sample stage 31, a bird-view camera (not shown) built into the stand arm 32 captures an overall observation image of the culture container 19.

Step S103: The CPU 42 instructs the observation unit 22 to take a microscopic image of the cells. The observation unit 22 lights up the LED light source 221 to illuminate the culture container 19, and drives the imaging device 224 to capture a microscopic image of the cells in the culture container 19. FIG. 5 shows an example of a microscope image captured by the imaging device 224.

FIG. 5 is a schematic diagram showing an example of a microscope image captured by the imaging device 224. In the example of FIG. 5, the microscope image PIC includes images of a cell Cell1, a cell Cell2, and a cell Cell3. This microscope image is a phase contrast image captured by a phase contrast microscope. This microscopic image shows the image of the cells cultured in the culture container 19 by converting the phase difference of the light transmitted through the cells for each part of the cell into a brightness difference of the light.

Returning to FIG. 4, the imaging device 224 captures a microscope image based on the imaging conditions (magnification of the objective lens, observation point in the container) specified by the user based on the management data stored in the storage unit 43. . For example, when observing a plurality of observation points within the culture container 19, the observation unit 22 sequentially adjusts the position of the culture container 19 by driving the sample stage 31, and captures a microscope image at each observation point. Note that the data of the microscopic image acquired in step S103 is read into the control device 41 (cultivation state evaluation device) and is recorded in the storage unit 43 under the control of the CPU 42.

Step S104: The CPU 42 instructs the container transport device 23 to transport the culture container 19 after the observation schedule ends. Then, the container transport device 23 transports the instructed culture container 19 from the sample stage 31 of the observation unit 22 to a predetermined storage position of the stocker 21. Thereafter, the CPU 42 ends the observation sequence and returns the process to step S101. This concludes the explanation of the flowchart in FIG. 4.

≪Culture condition evaluation processing in the culture condition evaluation device≫
Next, an example of the culture state evaluation process in the culture state evaluation device will be explained. In this example, an example will be described in which feature amounts of cells in the cultured cells in the culture container 19 are calculated using a plurality of microscope images obtained by time-lapse observation of the culture container 19.

In this example, the culture condition evaluation apparatus is provided with 24 culture vessels 19 in which three types of chemical substances are added at eight levels of concentration for one type of cell. That is, in this example, the culture condition evaluation device screens three types of chemical substances. Of the three types of chemical substances, one type is a chemical substance whose mechanism of action and other properties are unknown (hereinafter referred to as unknown chemical substances), and the second type is a chemical substance whose mechanism of action and other properties are known (hereinafter referred to as unknown chemical substances). (hereinafter referred to as known chemical substances).
In this embodiment, screening of chemical substances means selecting a chemical substance suitable for a purpose from among a plurality of chemical substances. Here, being fit for purpose means, for example, exhibiting specific medicinal efficacy or exhibiting specific toxicity. In this embodiment, as an example of screening of chemical substances, chemical substances are selected by determining the similarity between chemical substances. The chemical substances used for determining the similarity include, for example, unknown chemical substances and known chemical substances. That is, in the screening of chemical substances in this embodiment, as an example, unknown chemical substances are selected by determining the similarity between unknown chemical substances and known chemical substances.

In this culture state evaluation process, the control device 41 (culture state evaluation device) uses multivariate analysis to calculate the values calculated from the microscopic image in a space in which coordinate axes are defined for each type of a plurality of feature quantities related to the shape of cells. Points indicating the characteristic amount are determined for each concentration of the chemical substance added to the cell. Here, the feature amount is an amount indicating a feature related to the shape of a cell. Then, the control device 41 (culture state evaluation device) generates a graph made up of points indicating feature amounts in a space in which coordinate axes are defined for each type of feature amount related to the shape of the cell. The chemical substance screening method of this embodiment determines the similarity between different chemical substances based on a plurality of feature values related to the shape of cells included in microscopic images of cells to which different chemical substances have been added at different concentrations. have something to do.

≪Example of calculation model generation process≫
An example of the chemical substance screening operation will be described below with reference to the flowchart of FIG. In the example of FIG. 6, a culture vessel 19 in which a group of cells has been cultured is subjected to time-lapse observation of the same field of view under the same imaging conditions using the culture observation device 11, and a plurality of microscopic images of the cell group captured by time-lapse observation are prepared in advance. . In the following, a case where one type of cell is observed will be explained as an example. Twenty-four culture containers 19 are prepared for each type of cell. Chemical substances, which differ in type and concentration depending on the culture container 19, are added to the cells in the culture container 19. In the example of FIG. 6, in one time-lapse observation, one microscopic image is taken for each of the 24 culture vessels 19, each of which has a different type and concentration of chemical substances. Note that the number of microscopic images captured for each culture container 19 is not limited to one, but may be multiple. In the example of FIG. 6, 24 microscopic images are captured in one time-lapse observation. In the example of FIG. 6, the number of microscopic images to be captured is based on the number of sections of the culture container 19, the number of observation points (number of ROIs) by the culture observation device 11 set for each section, etc. May be set.
In other words, a microscopic image is a phase contrast image of cells to which different chemical substances have been added at different concentrations. Furthermore, there are multiple types of substances added to cells, and a microscopic image is an image taken for each type of substance. Microscopic images are taken at multiple times based on the time elapsed since the substance was added to the cells.
In addition, it is not necessary to use culture vessels 19 in which the types and concentrations of chemical substances to be added are different, and so that reproducibility can be obtained (confirmed) in determining the similarity of chemical substances. Culture vessels 19 having the same type and/or concentration may be used. In addition, the substance to be added to the cells may be of different types or concentrations, or may be added under different conditions.

In this example, the cell is any one type of cell among lung cancer cells, kidney cancer cells, skin cancer cells, prostate cancer cells, colon cancer, and brain tumors. In addition, in this embodiment, an example in which there is one type of cell will be described, but screening for chemical substances may be performed using a plurality of types of cells. The multiple types of cells may include normal cells as well as cancer cells.

Further, in this example, the chemical substances are three types of drugs: Paclitaxel, Etoposide, and unknown chemical substance X. Paclitaxel binds to cellular microtubules and inhibits their depolymerization. Etoposide inhibits the replication of cellular DNA. Chemical substance X is a target of chemical substance screening. That is, in this example, it is determined which of the known chemical substances Paclitaxel and Etoposide the unknown chemical substance X is most similar to.
In addition, in this example, the eight levels of concentration of the chemical substance are set such that the 50% inhibitory concentration (IC50) of each chemical substance is the median concentration, and the concentration ratio between adjacent concentrations is about 3 times. It is determined that Here, the 50% inhibitory concentration (IC50) is the concentration of a chemical substance at which the functions of half of the cells to which the chemical substance is added are inhibited.

Further, in the example of FIG. 6, time-lapse observation of the culture container 19 is performed for the first time after 8 hours have passed from the start of culture, and is performed every 8 hours until the 72nd hour. Therefore, in the example of FIG. 6, a set of nine microscopic images (8h, 16h, 24h, 32h, 40h, 48h, 56h, 64h, and 72h) of one culture container 19 is captured. That is, microscopic images are taken at multiple imaging times after the chemical is added to the cells. These captured microscopic images are stored in the storage unit 43 in advance. Hereinafter, the imaging time means the time at which a microscope image was captured in each time-lapse observation.

Step S201: The CPU 42 reads microscopic image data prepared in advance from the storage unit 43. In this example, the CPU 42 sequentially reads all microscope images prepared in advance as processing targets.

Step S202: The CPU 42 specifies an image to be processed from among the microscope images read in step S201. In this example, the CPU 42 sequentially designates all of the microscope images read in step S201 one by one as processing targets.

Step S203: The feature amount calculation unit 44 of the CPU 42 extracts images of cells included in the microscope image to be processed specified in step S202. For example, when a cell is imaged with a phase contrast microscope, a halo appears around areas where the phase difference changes significantly, such as the cell membrane. Therefore, the feature calculation unit 44 extracts a halo corresponding to a cell membrane using a known edge extraction method, and estimates that a closed space surrounded by edges is a cell through contour tracking processing. Thereby, the feature calculation unit 44 can extract images of individual cells from the microscope image. Note that the method by which the feature amount calculation unit 44 extracts images of individual cells from the microscope image may be a method other than the method using the edge extraction method described above. For example, the feature calculation unit 44 may extract images of individual cells from the microscope image using a known corner detection method.

Step S204: The feature amount calculation unit 44 calculates a plurality of feature amounts for each cell image extracted in step S203. Here, the plurality of feature amounts are a plurality of types of feature amounts related to the shape of cells included in microscopic images of cells to which different chemical substances have been added at different concentrations. Features related to the shape of cells include features related to the morphology of each cell, features related to the morphology of a group of cells, and features indicating changes over time in the features related to the morphology of a group of cells. included.
The types of feature amounts for the morphology of each cell include about 55 types of indicators such as cell area, length, width, length-to-width ratio, and perimeter. The characteristic amounts regarding the morphology of the cell group as a group include about 60 types of indicators such as the spread of cell distribution and the density of cells. Here, the spread of cell distribution is, for example, the dispersion of cell positions. As an example, the variance of the position of each cell is calculated based on the position of each cell in a predetermined direction of the microscope image. Approximately 600 types of indicators are used to indicate changes over time in features regarding the morphology of a group of cells, such as the rate of change over time in the spread of cell distribution and the rate of change in cell density over time. is included. The feature amount indicating the time change in the feature amount regarding the shape of the cell group as a group indicates the time change in the feature amount regarding the shape of the cell group as a group. Here, the time change in the feature amount is the time change in the feature amount for the time from the imaging time in the first time-lapse observation to the imaging time after the first time-lapse observation.

Step S205: The CPU 42 writes the feature amount of each cell calculated by the feature amount calculation unit 44 in step S204 into the storage unit 43. At this time, the CPU 42 associates the feature amount with the type of the feature amount, the type of chemical substance, the concentration of the chemical substance, and the imaging time, and writes the feature amount into the storage unit 43.

Step S206: The CPU 42 determines whether calculation of feature amounts has been completed for all the microscope images read from the storage unit 43 in step S201. If the CPU 42 determines that the calculation of feature amounts has been completed for all microscope images (step S206; YES), the process proceeds to step S207. If the CPU 42 determines that the calculation of the feature amount has not been completed (step S206; NO), the process returns to step S202.

Step S207: The coordinate calculation unit 45 of the CPU 42 calculates the feature amount space of the point specified by the set of values indicating each of the plurality of feature amounts related to the shape of the cell, based on the feature amount written in the storage unit 43 in step S205. Calculate the coordinates at . Here, the feature space is a space in which coordinate axes are defined for each type of feature amount related to the shape of a cell. The feature space is a space in which coordinate axes are defined for each type of feature amount related to the shape of a cell, so it is generally a high-dimensional space. The number of dimensions of the feature space is equal to the number of types of features related to the shape of the cell.

A point specified by a set of values indicating each of a plurality of feature quantities related to the shape of a cell is a representative value of each feature quantity regarding the morphology of each cell, and each representative value indicating a feature quantity regarding the morphology of a group of cells as a group. This is a point specified by a set of a value of , and each value representing a feature amount indicating a time change in the feature amount regarding the morphology of the cell group as a group.

Here, when calculating the coordinates, the coordinate calculation unit 45 calculates a representative value for the feature amount of the morphology of each cell, and calculates the coordinates of this representative value. The representative value of the feature amount is, for example, the average value or median value of the feature amount of each cell. Furthermore, the spread of cell distribution, which is one of the characteristic quantities regarding the morphology of a cell group as a group, includes the dispersion of the positions of each cell.
Hereinafter, a point specified by a set of values indicating each of a plurality of feature quantities related to the shape of a cell in the feature space will be referred to as a feature quantity point. Further, a graph made up of feature points is simply called a graph.

The coordinate calculation unit 45 calculates the coordinates of the feature points in the feature space for each microscope image having the same type of chemical substance, concentration of the chemical substance, and imaging time.
Here, there are three types of chemical substances: Paclitaxel, Etoposide, and Chemical Substance X, as described above. Furthermore, there are eight types of chemical substance concentrations determined such that the IC50 is the median value of the concentrations. Further, there are nine types of imaging times of the microscope images to be calculated: times every 8 hours (8h, 16h, 24h, 32h, 40h, 48h, 56h, 64h, and 72h).

The coordinate calculation unit 45 causes the storage unit 43 to store the coordinates of the feature points. Here, the process in step S207 of calculating the coordinates of the feature points in the feature space is an example of multivariate analysis. That is, the coordinate calculation unit 45 causes the storage unit 43 to store the results of the multivariate analysis.

Step S208: The coordinate calculation unit 45 of the CPU 42 performs principal component analysis on the feature points whose coordinates in the feature space were calculated in step S207.
The coordinate calculation unit 45 performs dimension reduction on the high-dimensional feature amount obtained from each cell by known principal component analysis, and uses it as principal component coordinates.
The feature points may be projected onto a low-dimensional space such as a two-dimensional or three-dimensional space for visualization. Visualization here means, for example, displaying feature points in the form of a graph. In the present embodiment, as an example, a case will be described in which the coordinate calculation unit 45 projects feature points onto a two-dimensional plane whose coordinate axes are axes PC1 and PC2 obtained as a result of principal component analysis. Hereinafter, a two-dimensional plane whose coordinate axes are axis PC1 and axis PC2 obtained as a result of principal component analysis will be referred to as a feature amount plane. Also, feature points projected onto a feature plane for visualization are called projected feature points.

FIG. 7 is a diagram showing an example of projected feature points processed by principal component analysis. That is, FIG. 7 is a diagram showing an example of projected feature points on the feature plane. In the graph shown in FIG. 7, the axis PC1 is an axis parallel to the direction in which the variation of feature points is greatest in the feature space. Further, in the graph shown in FIG. 7, the axis PC2 is an axis parallel to the direction in which the dispersion of feature points in the feature space is the largest among the directions orthogonal to the axis PC1. FIG. 7 shows a graph when three types of chemical substances, namely, chemical substance Ch1, chemical substance Ch2, and chemical substance Ch3 are added to a certain cell. In this example, the chemical substance Ch1 is Paclitaxel, the chemical substance Ch2 is Etoposide, and the chemical substance Ch3 is chemical substance X. Furthermore, in the graph shown in FIG. 7, plot P1-m indicates the projected feature value point of the cell to which the chemical substance Ch1 has been added. Furthermore, the subscript m of this plot P1-m indicates the concentration of the chemical substance added to the cells in order from the lowest to the highest. That is, the plot P1-1 shows the projected feature point of the cell to which the lowest concentration of chemical substance Ch1 was added. Plot P1-8 shows the projected feature point of the cell to which the highest concentration of chemical substance Ch1 was added.

Since the feature amounts in this embodiment include feature amounts indicating time changes in feature amounts regarding the morphology of a group of cells, the coordinate calculation unit 45 calculates the feature amount including time changes in the feature amounts. The principal component axes (axis PC1 and axis PC2) are determined from among them. Thereby, the coordinate calculation unit 45 can summarize many types of feature quantities into two-axis indicators without impairing information on the feature quantities including temporal changes. That is, according to the coordinate calculation unit 45, it is possible to visualize the morphological change of cells after adding a chemical substance while highlighting the characteristics of the morphological change.

Returning to FIG. 6, calculation of the density trajectory by the density trajectory calculation unit 46 will be explained.
Step S209: The density locus calculation unit 46 calculates the density locus of the feature point from the feature point whose coordinates in the feature space have been calculated by the coordinate calculation unit 45. Here, the concentration trajectory is a trajectory obtained by arranging feature points in order of concentration of chemical substances added to cells. In other words, the concentration trajectory calculation unit 46 calculates a trajectory formed by arranging the feature points calculated by the feature amount calculation unit 44 in the order of the concentration of the substance added to the cell.

Here, a case will be described in which the density locus calculation unit 46 calculates a density locus from feature points projected onto a two-dimensional plane by principal component analysis. A specific example of the density locus will be described with reference to FIG. 8. FIG. 8 is a diagram showing an example of a density trajectory according to this embodiment.
As shown in FIG. 8, the concentration locus calculation unit 46 calculates from a plot P1-1 of cells to which the lowest concentration of chemical substance Ch1 has been added to a plot P1-8 of cells to which the highest concentration of chemical substance Ch1 has been added. By arranging the above in order, the locus of the feature points is calculated. More specifically, the concentration trajectory calculation unit 46 connects the plot P1-1 and the plot P1-2 with a line segment L1. Further, the concentration trajectory calculation unit 46 connects the plot P1-2 and the plot P1-3 by a line segment L2. Furthermore, the concentration trajectory calculation unit 46 sequentially connects plots P1-3 to P1-8 by line segments L3 to L7.

Step S210: The display unit 50 displays the trajectory created in step S208 and step S209 as a graph. In this example, the display unit 50 displays the graph shown in FIG. 8.

Note that the process of projecting feature points onto a low-dimensional space of a two-dimensional or three-dimensional space and the process of displaying a trajectory as a graph are processes performed for visualization and may be omitted. That is, the processes of step S208, step S209, and step S210 may be omitted. In subsequent processing, processing is performed on feature points in the feature space. However, the subsequent processing may be performed on projected feature points projected onto a low-dimensional space such as a feature plane. For example, the subsequent processing may be performed on projected feature points on the feature plane shown in FIG. Even when processing is performed on projected feature points projected onto a low-dimensional space, the processing on projected feature points projected onto a low-dimensional space is the same as the processing on feature points in the feature space. It is.

<<Specific example of change area>>
Step S211: The changed region determining unit 47 determines a morphologically changed region based on the magnitude of change in the feature point in the feature space with respect to the change in the concentration of the chemical substance added to the cell. Here, the morphological change region is a range of concentrations of chemical substances in the feature space in which the magnitude of change in feature points with respect to changes in the concentration of chemical substances added to cells is greater than or equal to a predetermined value. In other words, the changed region determining unit 47 determines the morphologically changed region based on the change in the feature amount with respect to the change in the concentration of the chemical substance added to the cell.

Here, a feature point is a point specified by a set of representative values of each of a plurality of feature amounts related to the shape of a cell in the feature space, so a change in a feature point indicates a change in the shape of a cell. Further, the change in the feature value point with respect to the change in the concentration of the chemical substance added to the cell differs depending on the type of the chemical substance. For example, if a highly toxic chemical substance is added to a cell, even a slight increase in the concentration of the chemical substance will cause the cell to die, and in the process of dying, the cell will die due to changes in the concentration of the chemical substance. On the other hand, the feature values change rapidly. After the feature value changes rapidly (that is, after the cell dies), the cell remains dead and no change is observed in the feature value even if the concentration of the added chemical substance is increased. On the other hand, when a chemical substance that has a small effect on cells (for example, has low toxicity) is added to cells, no change is observed in the feature amount unless the concentration of the added chemical substance is increased to a certain level.

In this embodiment, the range of chemical substance concentrations is classified into a control region, a morphological change region, and a dead cell region.
The control region is a concentration range in which when the concentration of a chemical substance added to a cell is increased from the lowest concentration, the characteristic amount changes little with respect to the concentration of the chemical substance. In the control region, the effect of chemicals on cells is small. For chemical substances that have a small effect on cells, the range of the control region becomes larger.
The morphological change region is a concentration range in which the change in feature amount with respect to the concentration of a chemical substance becomes large when the concentration of a chemical substance added to a cell is increased. In the region of morphological change, the effects of chemical substances on cells are significant. Hereinafter, the concentration included in the shape change region will also be referred to as the shape change concentration. The morphological change concentration corresponds to the concentration at which the cell morphology changes.
The dead cell region is a concentration range in which, when the concentration of a chemical substance added to cells is increased, the cells die and the change in feature amount with respect to the concentration of the chemical substance becomes small. In the dead cell region, the cells are dead, so the feature amount becomes small, and therefore the change in the feature amount with respect to the concentration of the chemical substance also becomes small. Furthermore, when a chemical substance is added that is highly toxic to cells, the area of dead cells increases.

Note that in this embodiment, for convenience of explanation, the concentration of a chemical substance is expressed in units of predetermined concentration. For example, a density of 1 is a density that is 1 times the predetermined density, and a density of 8 is a density that is 8 times the predetermined density. In this embodiment, the lowest concentration among the concentrations of each chemical substance added to the cells is 1, and the highest concentration among the concentrations of each chemical substance added to the cells is 8.

The change region determining unit 47 determines a control region and a cell death region from a range of concentrations of a chemical substance added to cells, and forms a range of concentration of the chemical substance excluding the determined control region and cell death region. It is determined that the area is a changing area. Here, as an example, the changed region determination unit 47 determines a morphologically changed region based on a change in the feature amount with respect to a change in the concentration of a substance added to the cell. With reference to FIGS. 9 to 12, a method of determining a shape change area by the change area determination unit 47 will be described.

FIG. 9 is a diagram showing an example of the relationship between the distance of the feature point from the center of the control region and the chemical substance concentration according to the present embodiment. Here, the distance of the feature point from the center of the control area is the distance in the feature space. This distance is calculated by the change area determination unit 47 based on the coordinates of the feature point and the coordinates of the center of the control area. The unit of this distance is dimensionless. Graph Cd1 is a graph showing the relationship between the distance of the feature point from the center of the control region and the chemical substance concentration for Etoposide. Graph Cd2 is a graph showing the relationship between the distance of the feature point from the center of the control region and the chemical substance concentration for Paclitaxel. Graph Cd3 is a graph showing the relationship between the distance of the feature point from the center of the control area and the chemical substance concentration for chemical substance X.

In order to determine the control area, the changing area determination unit 47 first determines the coordinates of the center of the control area. Here, the center of the control region is, for example, the center of gravity between the feature points for the lowest concentration among the concentrations of each chemical substance added to the cells. For example, the center of the control region is the center of gravity of the feature point for Etoposide at a concentration of 1, the feature point for Paclitaxel at a concentration of 1, and the feature point for chemical substance X at a concentration of 1.

The change area determining unit 47 determines, in the feature space, feature points for each chemical substance whose distance from the center of the control region is smaller than the control distance, based on the coordinates of the feature points and the center of the control region. Determine based on coordinates. Here, the control distance is a predetermined distance common to each chemical substance. For example, the change area determining unit 47 determines a feature point for Etoposide whose distance from the center of the control area is smaller than the control distance. For example, the change area determining unit 47 determines a feature point for Paclitaxel whose distance from the center of the control area is smaller than the control distance. For example, the change region determination unit 47 determines a feature point for chemical substance X whose distance from the center of the control region is smaller than the control distance. Information indicating the control distance is stored in the storage unit 43.
Note that the control distance may be determined in advance for each chemical substance.

Note that the control distance may be a distance determined based on the coordinates of the feature point in the feature space with respect to the lowest concentration (concentration of 1) of each chemical substance added to the cell. For example, the control distance is determined such that the distance between the feature point and the control center for the lowest concentration (concentration of 1) of each chemical substance added to the cell is equal to or less than the control distance. Further, as the control distance, a distance determined in accordance with changes in cell morphology may be used. When a distance determined in accordance with cell morphological changes is used as the control distance, the larger the control distance, the greater the cell morphological changes relative to the concentration contained in the control region.
In this embodiment, the distance is, for example, the Euclidean distance or the Maharabinos distance.

The change region determining unit 47 determines that the concentration of the chemical substance corresponding to the determined feature point is a concentration included in the control region. The change area determination unit 47 determines a control area based on the determined density.

The change area determining unit 47 determines from the graph Cd1 that the distance of the feature point for Etoposide with chemical substance concentrations of 1 and 2 from the center of the control area is smaller than the control distance, the coordinates of the feature point and the center of the control area. Judgment is made based on the coordinates of Therefore, the change area determination unit 47 determines that the Etoposide control area has densities 1 and 2. From the graph Cd2, the change area determination unit 47 determines the coordinates of the feature points for Paclitaxel with chemical substance concentrations of 1, 2, 3, 4, and 5 if the distance from the center of the control area is smaller than the control radius. The determination is made based on the coordinates of the center of the control area. Therefore, the change area determination unit 47 determines that the control areas of Paclitaxel are concentrations 1, 2, 3, 4, and 5. The change area determination unit 47 determines from the graph Cd3 that the distances from the center of the control area of the feature points for chemical substance concentrations of 1, 2, 3, 4, 5, 6, and 7 for chemical substance X are smaller than the control radius. The determination is made based on the coordinates of the feature point and the coordinates of the center of the control area. Therefore, the change region determination unit 47 determines that the control region of chemical substance X has concentrations of 1, 2, 3, 4, 5, 6, and 7.

FIG. 10 is a diagram showing an example of the starting concentration of the shape change region for each chemical substance according to the present embodiment. The starting concentration of the morphological change region is the concentration of the chemical substance at which a plurality of types of feature amounts related to the shape of the cell begin to change by a predetermined amount. Here, the predetermined amount is the amount by which a plurality of types of feature amounts related to the shape of the cell change in the range of the concentration of the chemical substance corresponding to the control distance. For example, the starting concentration of the shape change region is the lowest concentration among the concentrations included in the shape change region. The change area determining unit 47 determines the next highest concentration of the chemical substance in the control area to be the starting concentration of the form change area. For example, the changed area determination unit 47 determines that the starting concentration of the morphological changed area of Etoposide is 3. The change area determination unit 47 determines that the starting concentration of the morphological change area of Cytochalasin B is 7. The change area determination unit 47 determines that the starting concentration of the form change area of the chemical substance X is 8.

In this manner, the changed area determination unit 47 determines the starting density of the morphologically changed area based on the control area. Here, as explained in FIG. 9, the control region is determined based on the change in the feature amount with respect to the change in the concentration of the chemical substance added to the cell. Therefore, the changed area determining unit 47 determines the morphological change concentration based on the change in the feature amount with respect to the change in the concentration of the chemical substance added to the cell. Further, as explained in FIG. 9, the morphological change density is determined by comparing the amount of change in the feature amount with a threshold value.

It can be seen that Etoposide, Doxorubicin, Latrunculin A, Vinblastine, and 17-AGG have lower starting concentrations in the morphological change region than other chemicals. It can be seen that 5-Fluorouracil, Cytochalasin B, and Paclitaxel have higher starting concentrations in the morphological change region than Etoposide, Doxorubicin, Latrunculin A, Vinblastine, and 17-AGG. It can be seen that chemical substance X has a higher initial concentration in the shape change region than other chemical substances.

FIG. 11 is a diagram showing an example of the relationship between the distance of the feature point from the center of the dead cell region and the chemical substance concentration according to the present embodiment. Here, the distance of the feature point from the center of the dead cell region is the distance in the feature space. This distance is calculated by the changed area determining unit 47 based on the coordinates of the feature point and the coordinates of the center of the dead cell area. Graph Dd1 is a graph showing the relationship between the distance of the feature point from the center of the dead cell region and the chemical substance concentration for Etoposide. Graph Dd2 is a graph showing the relationship between the distance of the feature point from the center of the dead cell region and the chemical substance concentration for Paclitaxel. Graph Dd3 is a graph showing the relationship between the distance of the feature point from the center of the dead cell region and the chemical substance concentration for chemical substance X.

The changed region determination unit 47 determines, in the feature space, feature points for each chemical substance whose distance from the center of the dead cell region is smaller than the dead cell distance as feature points included in the dead cell region. It is determined that Here, the center of the dead cell area is, for example, the center of gravity between the feature points for the highest concentration of each chemical substance added to the cell. The center of the dead cell region is, for example, the center of gravity of the feature point for Etoposide at a concentration of 8, the feature point for Paclitaxel at a concentration of 8, and the feature point for chemical substance X at a concentration of 8. The dead cell distance is a predetermined distance common to chemical substances. For example, the changed region determining unit 47 determines a feature point for Etoposide whose distance from the center of the dead cell region is smaller than the dead cell distance. For example, the changed region determining unit 47 determines a feature point for Paclitaxel whose distance from the center of the dead cell region is smaller than the dead cell distance. For example, the changed region determination unit 47 determines a feature point for chemical substance X whose distance from the center of the dead cell region is smaller than the dead cell distance. Information indicating the dead cell distance is stored in the storage unit 43.
Note that the dead cell distance may be determined in advance for each chemical substance.

Note that, as the dead cell distance, a distance determined based on the coordinates in the feature space of the feature point for the highest concentration (concentration 8) of each chemical substance added to the cell may be used. Further, as the dead cell distance, a distance determined based on changes in cell morphology may be used. When a distance determined in accordance with changes in cell morphology is used as the dead cell distance, the larger the dead cell distance is, the greater the change in cell morphology with respect to the concentration contained in the dead cell region.

From the graph Dd1, the change area determining unit 47 determines that the distance of the feature point for chemical substance concentrations 7 and 8 for Etoposide from the center of the dead cell region is smaller than the dead cell distance, the coordinates of the feature plot and the dead cell. Determine based on the coordinates of the center of the area. Therefore, the changed region determining unit 47 determines that the dead cell regions of Etoposide have concentrations of 7 and 8. The change area determining unit 47 determines from the graph Dd2 that the distances of the feature points for chemical substance concentrations of 6, 7, and 8 for Paclitaxel from the center of the dead cell region are smaller than the dead cell distance, the coordinates of the feature points, Judgment is made based on the coordinates of the center of the dead cell area. Therefore, the changed region determining unit 47 determines that the dead cell regions of Paclitaxel have concentrations of 6, 7, and 8. The change area determining unit 47 determines from the graph Dd3 that there is no chemical substance concentration for chemical substance X where the distance from the center of the dead cell area is smaller than the dead cell distance, and determines the coordinates of the feature point and the dead cell area. Judgment is made based on the coordinates of the center of. Therefore, the changed region determining unit 47 determines that there is no dead cell region of chemical substance X.

FIG. 12 is a diagram showing an example of dead cell areas for each chemical substance according to this embodiment. For each chemical substance, the changed area determining unit 47 determines, for each chemical substance, a concentration range that is higher than the starting concentration of the morphologically changed area, excluding the dead cell area, as the morphologically changed area. In other words, the change region determination unit 47 determines the shape of the shape change region by excluding the concentration range in which the change in a plurality of types of feature quantities related to the shape of the cell is equal to or less than a predetermined amount from the concentration range higher than the starting concentration of the shape change region. Determined as a changing area. For example, the changed region determining unit 47 determines, for Etoposide, a region obtained by excluding the range of concentrations 7 and 8, which are dead cell regions, from the range of concentrations higher than the starting concentration 3 of the morphologically changed region, as the morphologically changed region. For example, the changed region determination unit 47 determines that the concentration range of Paclitaxel is determined by excluding the concentration range of concentrations 6, 7, and 8, which are dead cell regions, from the concentration range higher than concentration 6, which is the starting concentration of the morphological change region. Since no region remains, it is determined that no shape change region exists. Since there is no dead cell region for the chemical substance X, the changed region determination unit 47 determines the concentration range of 8, which is a concentration range higher than the starting concentration of the morphological change region, to be the morphological change region.

Thus, the morphological change concentration includes the concentration from when the cell's morphology changes until the cell dies.
In addition, in this way, the changed area determination unit 47 determines the morphologically changed area based on the dead cell area. Here, as explained in FIG. 11, the dead cell area is determined based on the change in the feature amount with respect to the change in the concentration of the chemical substance added to the cell. Therefore, the changed area determining unit 47 determines the morphological change concentration based on the change in the feature amount with respect to the change in the concentration of the chemical substance added to the cell. Further, as explained in FIG. 9, the morphological change density is determined by comparing the amount of change in the feature amount with a threshold value.

Returning to FIG. 6 again, the explanation of the chemical substance screening operation will be continued.
Step S212: The similarity determination section 48 determines the similarity between chemical substances by comparing the morphological change regions determined by the change region determination section 47 in step S211 between the chemical substances added to the cells. do. That is, the similarity determining unit 48 compares the feature amount corresponding to the morphological change concentration at which the cell morphology changes, among different feature amounts, between different chemical substances, and determines the similarity between the chemical substances. Therefore, the similarity determining unit 48 determines the similarity between chemical substances based on the change in the feature amount with respect to the change in concentration indicated by the results of the multivariate analysis in steps S207 and S211. In other words, the similarity determination unit 48 determines the similarity between chemical substances based on the results of multivariate analysis of a plurality of feature amounts.

Here, similarity between chemical substances means that when a chemical substance is added to a cell, the types and magnitude of changes in the features of the cell shape that the chemical substance changes are similar between the chemical substances. That's true. In this embodiment, by determining the similarity between chemical substances, it is possible to determine the similarity in properties such as the mechanism of action between the chemical substances. That is, since it is possible to determine the similarity of properties such as the mechanism of action of an unknown chemical substance to a known chemical substance, it is possible to efficiently screen chemical substances.

Now, with reference to FIG. 13, details of the process in which the similarity determining unit 48 determines the similarity between chemical substances in step S212 in FIG. 12 will be described.
FIG. 13 is a flowchart illustrating an example of the operation of comparing feature points in the feature space between chemical substances by the control device according to the present embodiment.

Step S301: Regarding the feature points in the feature space of the target chemical substance for which similarity is to be determined, for each feature point included in one shape change region, the similarity determination unit 48 determines whether the feature points included in one shape change region are Among the included feature points, the closest feature point is searched based on the coordinates of the feature point.

Here, the operation of the similarity determination unit 48 in step S301 will be explained with reference to FIG. 14. FIG. 14 is a diagram illustrating an example of a method of comparing feature points in the feature space between chemical substances by the control device according to the present embodiment. The processing of the similarity determination unit 48 is performed on feature points in a high-dimensional feature space. However, in FIG. 14, projected feature points for chemical substance A, chemical substance B, and chemical substance X are shown on the feature plane obtained as a result of principal component analysis for visualization. In other words, the feature plane shown in FIG. 14 is a plane from which a certain plane of the feature space is extracted, and the projected feature points shown in FIG. It is a point.

In the example shown in FIG. 14, the similarity determination unit 48 determines the similarity of chemical substance X to chemical substance A and the similarity to chemical substance B.
The morphological change area ChA for chemical substance A consists of plots PA-1 to PA-5. The morphological change region ChB for chemical substance B consists of plots PB-1 to PB-5. The morphological change region ChX for chemical substance X consists of plots PX-1 to PX-3.

The similarity determination unit 48 searches for the feature point having the closest distance from the plot PX-1 included in the shape change area ChX from among the plots PA-1 to PA-5 included in the shape change area ChA. . The similarity determining unit 48 determines that the plot PA-3 whose distance from the plot PX-1 is the distance d1 is the feature point whose distance from the plot PX-1 is the closest, based on the coordinates of the feature point. do. Similarly, the similarity determination unit 48 determines that the plot PA-4 whose distance from the plot PX-2 is the distance d2 is the feature point whose distance from the plot PX-2 is the closest. Judgment based on The similarity determining unit 48 determines that the plot PA-5 whose distance from the plot PX-3 is the distance d3 is the feature point whose distance from the plot PX-3 is the closest, based on the coordinates of the feature point. do.
However, it is assumed that the concentration of the chemical substance corresponding to the feature point included in the morphological change region is the same as the concentration of the chemical substance corresponding to the feature point determined to be closest in distance from this feature point. is not limited. For example, the concentration of a chemical substance corresponding to plot PX-1 and the concentration of a chemical substance corresponding to plot PA-3 are not necessarily the same.

The similarity determination unit 48 searches for the feature point having the closest distance from the plot PX-1 included in the morphological change area ChX from among the plots PB-1 to PB-5 included in the morphological change area ChB. . The similarity determination unit 48 determines that the plot PB-2 whose distance from the plot PX-1 is the distance d4 is the feature point whose distance from the plot PX-1 is the closest based on the coordinates of the feature point. do. Similarly, the similarity determination unit 48 determines that the plot PB-3 whose distance from the plot PX-2 is the distance d5 is the feature point whose distance from the plot PX-2 is the closest. Judgment based on The similarity determination unit 48 determines that the plot PB-3 whose distance from the plot PX-3 is a distance d6 is the feature point whose distance from the plot PX-3 is the closest based on the coordinates of the feature point. do.

Returning to FIG. 13, the operation of comparing feature points on the feature plane between chemical substances by the control device will be continued.
Step S302: The similarity determination unit 48 calculates the average distance between the feature point of the shape change region of chemical substance X and the closest feature point of the shape change region of chemical substance A determined in step S301. Calculate based on the coordinates of the quantitative point. The similarity determination unit 48 calculates, for example, the average of three distances, distance d1, distance d2, and distance d3, between the shape change area ChX and the shape change area ChA based on the coordinates of the feature points.
On the other hand, the similarity determination unit 48 calculates the average distance between the feature point of the shape change region of chemical substance X and the closest feature point of the shape change region of chemical substance B determined in step S301 as a feature value. Calculate based on the coordinates of the point. The similarity determination unit 48 calculates, for example, the average of three distances, distance d4, distance d5, and distance d6, between the shape change region ChX and the shape change region ChB based on the coordinates of the feature points.

Step S303: The similarity determination unit 48 determines the known chemical substance that is most similar to the unknown chemical substance X from among the chemical substances that are the targets of similarity determination. The similarity determining unit 48 determines that the chemical substance with the smallest average distance between the feature amount points of the shape change region calculated in step S302 is the known chemical substance that is most similar to the unknown chemical substance X.
In the example shown in FIG. 14, the similarity determining unit 48 determines that the chemical substance A having the morphological change region ChA is the known chemical substance that is most similar to the chemical substance X.

Based on the determination result, the control device 41 can classify the unknown chemical substance X with respect to the morphological change caused in cells.
In the present embodiment, a case has been described in which the control device 41 determines the similarity between the unknown chemical substance X and the known chemical substance, but the present invention is not limited to this. The control device 41 may determine the similarity between known chemical substances.

≪Summary≫
As explained above, the chemical substance screening method provided by the control device 41 of the present embodiment uses a plurality of feature quantities related to the shape of cells included in microscopic images of cells to which different chemical substances have been added at different concentrations. The method includes determining the similarity between different chemical substances based on the same, and outputting the determined result.
With this configuration, in the chemical substance screening method provided by the control device 41 of the present embodiment, the similarity between different chemical substances can be determined based on a plurality of feature amounts related to the shape of cells, so that the chemical substance screening method is cost effective. and time can be reduced. In the chemical substance screening method provided by the control device 41 of the present embodiment, by determining the similarity between chemical substances, it is possible to determine the similarity in properties such as the mechanism of action between the chemical substances. That is, in the chemical substance screening method provided by the control device 41 of the present embodiment, it is possible to determine the similarity of properties such as the action mechanism of an unknown chemical substance to a known chemical substance, so that chemical substance screening can be performed efficiently. It can be performed.

Furthermore, in the chemical substance screening method provided by the control device 41 of this embodiment, the similarity between chemical substances is determined based on the results of analyzing a plurality of feature amounts by multivariate analysis.
With this configuration, in the chemical substance screening method provided by the control device 41 of the present embodiment, the similarity between chemical substances can be determined based on the results of multivariate analysis of a plurality of feature amounts. , the accuracy of chemical substance screening can be increased compared to the case where similarity between chemical substances is not determined based on the results of multivariate analysis.

Furthermore, in the chemical substance screening method provided by the control device 41 of the present embodiment, determining the similarity between different chemical substances corresponds to the morphological change concentration at which the cell morphology changes among the different feature quantities. The feature values of different chemical substances are compared to determine the similarity between different chemical substances.
With this configuration, in the chemical substance screening method provided by the control device 41 of the present embodiment, similarity between chemical substances is determined by limiting the concentration range of the chemical substance to which the chemical substance causes a large morphological change in cells. This reduces the cost and time of chemical screening compared to determining similarity between chemicals without limiting the range of concentrations of chemicals that cause large morphological changes in cells. be able to.

Further, the chemical substance screening method provided by the control device 41 of the present embodiment includes determining the morphological change concentration based on the change in the feature amount with respect to the change in the concentration of the chemical substance added to the cell.
With this configuration, in the chemical substance screening method provided by the control device 41 of the present embodiment, similarity between chemical substances can be determined limited to a concentration range higher than the concentration at which cell morphology changes. , the cost and time in screening chemicals can be reduced compared to not limiting the range of concentrations above the concentration at which cell morphology changes begin.

In addition, in the chemical substance screening method provided by the control device 41 of the present embodiment, the morphological change concentration includes the concentration from when the cell morphology changes until the cell dies.
With this configuration, in the chemical substance screening method provided by the control device 41 of the present embodiment, similarity between chemical substances can be determined limited to a concentration range lower than the concentration that causes cell death. The cost and time of chemical screening can be reduced compared to not limiting the range of concentrations below those that cause cell death.

In the chemical substance screening method provided by the control device 41 of the present embodiment, determining the morphological change concentration involves comparing the amount of change in the feature amount with a threshold value to determine the morphological change concentration.
With this configuration, in the chemical substance screening method provided by the control device 41 of the present embodiment, the morphological change concentration is obtained by comparing the amount of change in the feature amount with the threshold value, so that the morphological change concentration can be easily obtained. can.

Further, in the chemical substance screening method provided by the control device 41 of the present embodiment, the different chemical substances include known chemical substances and unknown chemical substances, and the known chemical substances and unknown chemical substances. Determine the similarity of.
With this configuration, it is possible to determine the similarity between a known chemical substance and an unknown chemical substance, so it is possible to determine which of the known chemical substances the unknown chemical substance is similar to.

Note that the control device 41 performs the process of analyzing the plurality of feature quantities by multivariate analysis in step S207 and step S211, instead of performing the process of analyzing the plurality of feature quantities by multivariate analysis in step S207 and step S211. may be performed by an external device. Here, the external device is a device provided outside the control device 41. This external device may be provided in the lower casing 13 separately from the control device 41, or may be provided independently from the culture observation device 11. This external device includes a processor that executes arithmetic processing of multivariate analysis, and a storage unit that stores analysis results of multivariate analysis.

When the control device 41 causes an external device to perform the process of analyzing a plurality of feature quantities by multivariate analysis in steps S207 and S211, the CPU 42 in step S205 analyzes each cell calculated by the feature quantity calculation unit 44 in step S204. The feature quantity is supplied to an external device. In step S<b>212 , the CPU 42 acquires the results of the process performed by the external device to analyze a plurality of feature amounts by multivariate analysis, and supplies the results to the similarity determination unit 48 . The similarity determination unit 48 determines the similarity between different chemical substances based on the results of a process performed by an external device in which a plurality of feature amounts are analyzed by multivariate analysis.

Further, the control device 41 may cause an external device to perform the process of calculating and storing a plurality of feature amounts related to the shape of cells included in the microscope image from step S201 to step S205. This external device may be the same device as the above-mentioned external device, or may be a different device from the above-mentioned external device. When the control device 41 causes an external device to perform the processes from step S201 to step S205, the control device 41 acquires a plurality of feature amounts calculated by the external device and performs the processes from step S206 onwards.

The display unit 50 displays the determination result of the similarity between different chemical substances determined by the similarity determination unit 48. In other words, the chemical substance screening method of this embodiment includes outputting the determined results.

In the present embodiment, a case has been described in which the changed region determining section 47 determines a control region and a cell death region when determining a morphologically changed region in step S211, but the present invention is not limited to this. Information indicating the control region and information indicating the cell death region may be stored in the storage unit 43 in advance. If the information indicating the control area and the information indicating the cell death area are stored in the storage unit 43 in advance, the changed area determination unit 47 receives the information indicating the control area and the information indicating the cell death area from the storage unit 43. The range of concentration of the chemical substance excluding the control region indicated by the obtained information indicating the control region and the cell death region indicated by the obtained information indicating the cell death region may be determined to be the morphological change region.

Further, in the present embodiment, in step S211, the changed region determining unit 47 determines the starting concentrations of the control region, the cell death region, and the morphologically changed region in order to determine the morphologically changed region, but the invention is not limited to this. . In step S211, the changed area determination unit 47 may determine at least one of the control area, the cell death area, and the starting concentration of the morphological change area, instead of determining the morphological change area. In other words, the change region determining unit 47 may determine the concentration at which cell morphology starts to change based on the results of multivariate analysis. Further, the changed region determination unit 47 may determine the concentration that causes cell death of the previous cell based on the results of multivariate analysis.

Further, the changed region determination unit 47 may determine the morphologically changed region without determining the control region and the cell death region. For example, the changed area determination unit 47 may determine that the density included in the upper and lower limits of the distance is the morphological changed area. The change region determination unit 47 may determine the concentration at which the change in cell morphology begins and the concentration at which the change in cell morphology ends based on the results of multivariate analysis. The changing region determination unit 47 may determine the range of concentrations from the determined concentration at which the change in cell morphology begins to the concentration at which the change in cell morphology ends to be the morphological change region.

When the changed region determination unit 47 determines at least one of the control region, cell death region, and morphological change region starting concentration in step S211, the similarity determination unit 48 determines the control region, cell death region, and morphological change region in step S212. The concentration range of the chemical substance may be limited based on at least one of the starting concentration of the region and the shape change region, and the similarity between the chemical substances may be determined.

For example, the similarity determining unit 48 may determine the similarity between chemical substances by comparing the concentration range of chemical substances excluding the control region between the chemical substances added to the cells. In another example, the similarity determining unit 48 determines the similarity between chemical substances by comparing the concentration range of the chemical substances excluding the cell death area between the chemical substances added to the cells. It's fine. In another example, the similarity determination unit 48 compares the concentration range of chemical substances higher than the starting concentration of the morphological change region between the chemical substances added to the cells, thereby determining the similarity between the chemical substances. You can determine the gender. Further, the similarity determination unit 48 may determine the similarity between chemical substances in the control region. The similarity determining unit 48 may determine the similarity between chemical substances in the cell death region.

In the above embodiment, the control region is determined based on a predetermined distance from the center of gravity between feature points for the lowest concentration of each chemical substance added to cells, and the dead cell region is We have explained the case where the determination is made based on the predetermined distance from the center of gravity between the feature points for the highest concentration of each added chemical substance, but the control area and dead cell area are determined based on the known chemical substance obtained in advance. may be determined using data from

Note that in the above embodiment, a case has been described in which the known chemical substances are Paclitaxel and Etoposide, but the present invention is not limited thereto. A known chemical can be any chemical that exists. For example, known chemicals may be PD-180970, Doxorubicin, 5-Fluorouracil, Cytochalasin B, Latrunculin A, Vinblastine, and 17-AAG shown in FIGS. 10 and 12.

Further, in the above embodiment, a case has been described in which there are two types of known chemical substances, but the present invention is not limited to this. In screening for chemical substances in the above embodiments, one type, or three or more types of known chemical substances may be used.

Furthermore, in the above embodiment, when determining the control region, the center of the control region is the center of gravity between the feature points for the lowest concentration among the concentrations of each chemical substance added to the cell. Although explained above, it is not limited to this. The center of the control region may be determined based on changes in cell morphology. That is, the center of the control region may be determined based on a change in the feature amount.

In addition, in the above embodiment, when determining a dead cell area, the center of the dead cell area is, for example, between the feature points for the highest concentration of each chemical substance added to the cell. Although the case where the center of gravity is the center of gravity has been described, the present invention is not limited to this. The center of the dead cell area may be determined, for example, based on the ratio of dead cells to live cells. If the center of the dead cell area is determined based on the ratio of dead cells to living cells, the center of the dead cell area is determined based on the concentration of the chemical substance at which the ratio of dead cells to living cells is greater than or equal to a predetermined ratio. Determined based on quantity points.

Further, in the above embodiment, as an example, a case has been described in which the similarity between different chemical substances in the morphological change region is determined, but the present invention is not limited to this. For example, similarity may be determined over the entire concentration range of added chemicals. Further, the similarity may be determined within a certain concentration range of the added chemical substance.

(Second embodiment)
A second embodiment of the present invention will be described in detail below with reference to the drawings.
In the first embodiment, the control device determines the concentration of a chemical substance that causes a large morphological change in cells, and then determines the similarity between the chemical substances. In this embodiment, a case will be described in which the control device determines the similarity between chemical substances based on the temporal change in cell morphology when a chemical substance at a 50% inhibitory concentration (IC50) is added.
The culture observation device used in this embodiment is called a culture observation device 11a.

FIG. 15 is a block diagram showing an overview of the culture observation device 11a used in this embodiment. Comparing the culture observation device 11a (FIG. 15) according to this embodiment with the culture observation device 11 (FIG. 1) according to the first embodiment, the control device 41a is different. However, the functions of the other components (upper casing 12, lower casing 13, observation unit 22, and display section 50) are the same as in the first embodiment. Description of the same functions as in the first embodiment will be omitted, and in the second embodiment, the description will focus on parts that are different from the first embodiment.

The control device 41a includes a CPU 42a and a storage section 43. The CPU 42a is a processor that executes various calculation processes of the control device 41a. The CPU 42a is configured of an integrated circuit such as an LSI (Large Scale Integration), for example. Note that part or all of the CPU 42a may be configured as software. That is, the CPU 42a may be configured by combining hardware and software. In addition, the CPU 42 functions as a feature value calculation section 44a, a coordinate calculation section 45a, a time trajectory calculation section 46a, a time trajectory quantification section 47a, and a similar chemical substance determination section 48a, respectively, by executing the program.

≪Example of chemical substance screening operation≫
An example of the chemical substance screening operation will be described below with reference to the flowchart of FIG. 16. FIG. 16 is a flowchart showing an example of the operation of the analysis of known chemical substances and the analysis of unknown chemical substances by the control device 41a used in this embodiment.

Step S401: The control device 41a analyzes known chemical substances. Here, the process of step S401 is performed after the same process as the process from step S201 to step S209 in FIG. 6 has been performed. The feature amount calculation unit 44a calculates feature amounts for each extracted cell image. The feature amount calculation unit 44a calculates a feature amount for each microscope image corresponding to the elapsed time when the cells were imaged. The types of feature amounts calculated by the feature amount calculation unit 44a include feature amounts regarding the morphology of each cell and feature amounts regarding the morphology of a group of cells.
The types of feature amounts for the morphology of each cell include about 55 types of indicators such as cell area, length, width, length-to-width ratio, and perimeter. The characteristic amounts regarding the morphology of the cell group as a group include about 60 types of indicators such as the spread of cell distribution and the density of cells.

The coordinate calculation unit 45a calculates features for each type of feature, type of chemical substance, concentration of chemical substance, and imaging time of the microscope image to be calculated, which are associated with the feature calculated by the feature calculation unit 44a. Calculate the coordinates of the feature points in the quantity space.

Here, the time trajectory calculation unit 46a may calculate the time trajectory and density trajectory of the feature point from the feature point whose coordinates have been calculated by the coordinate calculation unit 45a. Here, the time trajectory refers to the feature points in the feature space that connect the feature points for each concentration of the chemical substance added to the cells using line segments or curves in the order of the imaging time of the microscope image. This is the trajectory obtained by That is, the time trajectory calculation unit 46a calculates a time trajectory, which is a trajectory formed by arranging the feature points calculated by the feature amount calculation unit 44a in the order of elapsed time when the cells were imaged. In addition, a concentration trajectory is a line or curve that connects feature points in the feature space for each imaging time of a microscope image in the order of the concentration of the chemical substance added to the cell. This is the trajectory obtained by

Here, with reference to FIG. 17, specific examples of the concentration trajectory and the time trajectory will be described. The concentration trajectory and the time trajectory are calculated in a high-dimensional feature space. However, for visualization purposes, FIG. 17 shows an example in which feature points in a high-dimensional feature space are projected onto projected feature points on a feature plane by principal component analysis. Similarly, FIGS. 18, 19, and 21, which will be described later, show examples in which feature points in a high-dimensional feature space are projected onto projected feature points on a feature plane by principal component analysis.
17, FIG. 18, FIG. 19, and FIG. 21 show an example in which the time trajectory calculation unit 46a calculates the time trajectory and density trajectory of the feature point, but the time trajectory calculation unit 46a It is not necessary to calculate the time trajectory and concentration trajectory. That is, in FIGS. 17, 18, 19, and 21, the feature points do not need to be connected using line segments or curves.

FIG. 17 is a diagram showing an example of a concentration trajectory on the high concentration side and a time trajectory at a 50% inhibitory concentration according to the present embodiment. FIG. 17 shows the time trajectory when a chemical substance is added to a certain cell at an 50% inhibitory concentration (IC50). This time trajectory is shown on a two-dimensional plane (feature plane) whose coordinate axes are the axis PC1 and the axis PC2 obtained as a result of principal component analysis. This time trajectory consists of the plot IC50-t. However, the subscript t of the plot IC50-t indicates the imaging time of the microscope image in chronological order. That is, the plot IC50-1 shows the projected feature amount point corresponding to the concentration of 50% inhibition concentration (IC50) for the oldest imaging time. Further, the plot IC50-5 shows the projected feature point corresponding to the concentration of 50% inhibition concentration (IC50) for the latest imaging time.

FIG. 17 shows high concentration plots PH1 to PH3, which are projected feature points corresponding to concentrations higher than the 50% inhibitory concentration (IC50). The imaging time of the microscope images corresponding to the high density plots PH1 to PH3 is the same imaging time as the imaging time of the microscope image corresponding to the plot IC50-3. In the high concentration plots PH1 to PH3, the concentrations of the corresponding chemical substances increase in the order of high concentration plot PH1, high concentration plot PH2, and high concentration plot PH3.
For the high concentration plot PH1 to high concentration plot PH3, the imaging time of the microscope image corresponding to the plot IC50-1, plot IC50-2, plot IC50-4, and plot IC50-5 is the same as for plot IC50-3. A time trajectory (high concentration plot PH1-t to high concentration plot PH3-t) corresponding to is calculated.
Plot IC50-3, high concentration plot PH1, high concentration plot PH2, and high concentration plot PH3 are connected by high concentration side line segment LH1 to high concentration side line segment LH3 to form a concentration locus.

FIG. 18 is a diagram showing an example of a concentration trajectory on the low concentration side and a time trajectory at a 50% inhibitory concentration according to the present embodiment. In addition to the time trajectory (plot IC50-t) shown in FIG. 17, FIG. 18 also shows low concentration plots PL1 to PL3, which are plots corresponding to concentrations lower than the 50% inhibitory concentration (IC50). It is shown. The low concentration plots PL1 to PL3 are shown on a two-dimensional plane (feature plane) whose coordinate axes are the axis PC1 and the axis PC2 obtained as a result of principal component analysis. The imaging time of the microscope images corresponding to the low density plots PL1 to PL3 is the same imaging time as the imaging time of the microscope image corresponding to the plot IC50-3. In the low concentration plots PL1 to PL3, the concentrations of the corresponding chemical substances decrease in the order of low concentration plot PL1, low concentration plot PL2, and low concentration plot PL3.
For low concentration plots PL1 to PL3, similarly to plot IC50-3, the imaging time of the microscope images corresponding to plot IC50-1, plot IC50-2, plot IC50-4, and plot IC50-5 A time trajectory (low concentration plot PL1-t to low concentration plot PL3-t) corresponding to is calculated.
The plot IC50-3, the low concentration plot PL1, the low concentration plot PL2, and the low concentration plot PL3 are connected by the low concentration side line segment LL1 to the low concentration side line segment LL3 to form a concentration locus.

Returning to FIG. 16, the explanation of an example of the chemical substance screening operation will be continued.
Step S402: The time trajectory quantification unit 47a quantifies the feature points in the feature space calculated by the coordinate calculation unit 45a for each concentration of the chemical substance added to the cells. Furthermore, the time trajectory quantification unit 47a quantifies the feature points in the feature space calculated by the coordinate calculation unit 45a for each imaging time of the microscope image. Here, the quantification of feature points for each concentration of a chemical substance added to cells is, for example, calculation of the distance between feature points at a 50% inhibitory concentration. Quantification of feature points for each imaging time of a microscope image includes, for example, calculating the sum of distances between feature points at concentrations higher than 50% inhibitory concentration, and calculating features at concentrations lower than 50% inhibitory concentration. This is a calculation of the sum of distances between quantitative points. The sum of the distances between the feature points on the higher concentration side than the 50% inhibitory concentration and the sum of the distances between the feature points on the lower concentration side than the 50% inhibitory concentration are calculated at each imaging time of the microscope image. Ru. The distance between feature points is, for example, Euclidean distance or Maharabinos distance.

FIG. 19 is a diagram showing an example of a time trajectory at a 50% inhibitory concentration of the first chemical substance according to the present embodiment. This time trajectory is shown on a two-dimensional plane (feature plane) whose coordinate axes are the axis PC1 and the axis PC2 obtained as a result of principal component analysis. The first chemical substance is, for example, Paclitaxel. The time trajectory at the 50% inhibitory concentration of the first chemical substance consists of five feature points, plots PT1 to PT5.

FIG. 20 is a diagram showing an example of a temporal change in the sum of distances between feature points on the low concentration side and the sum of distances between feature points on the high concentration side of the first chemical substance according to the present embodiment. be. In FIG. 20, the sum of the line segments on the low-density side and the sum of the line segments on the high-density side are shown with respect to the imaging time. The time trajectory quantification unit 47a calculates the sum of low-density line segments SL1 and the sum of high-density line segments SH1 at each of five times t1 to t5.

FIG. 21 is a diagram showing an example of a time trajectory at a 50% inhibitory concentration of the second chemical substance according to the present embodiment. In FIG. 21, the time trajectory is shown on a two-dimensional plane (feature plane) whose coordinate axes are the axis PC1 and the axis PC2 obtained as a result of principal component analysis. The second chemical substance is, for example, Etoposide. The time trajectory at 50% inhibitory concentration of the second chemical consists of five plots, plots ET1 to ET5.

FIG. 22 is a diagram showing an example of a temporal change in the sum of distances between feature points on the low concentration side and the sum of distances between feature points on the high concentration side of the second chemical substance according to the present embodiment. be. In FIG. 22, the sum of the distances between the feature points on the low density side and the sum of the distances between the feature points on the high density side are shown with respect to the imaging time. The time trajectory quantification unit 47a calculates the low-density side line segment sum SL2 and the high-density side line segment sum SH2 at each of five times t1 to t5.

Returning to FIG. 16, the explanation of an example of the chemical substance screening operation will be continued.
Step S403: The time trajectory quantification unit 47a registers the quantified data calculated in step S402 in the database. The quantified data calculated in step S402 are the sum of distances between feature points at 50% inhibitory concentration, the sum of distances between feature points at concentrations higher than 50% inhibitory concentration, and the 50% inhibitory concentration. This is the sum of distances between feature points on the lower density side. The database is, for example, the storage unit 43.

Step S404: The control device 41a analyzes the unknown chemical substance in the same way as it did for the known chemical substance in step S401. Specifically, the control device 41a plots feature points in the feature space based on the distribution of the features for the unknown chemical substance. The control device 41 may also calculate a time trajectory and a concentration trajectory obtained from the analysis results.
Step S405: The time trajectory quantification unit 47a calculates the feature point in the feature space calculated by the coordinate calculation unit 45a for the unknown chemical substance, in the same way as it did for the known chemical substance in step S402. The feature points are quantified for each concentration of chemical substances added.

Step S406: The similar chemical substance determining unit 48a determines which of the known chemical substances the unknown chemical substance is most similar to. In this embodiment, the similar chemical substance determining unit 48a determines whether the unknown chemical substance is a first chemical substance (Paclitaxel) which is a known chemical substance and a second chemical substance (Etoposide) which is a known chemical substance. A judgment is made as to which of the following is most similar. In other words, in the chemical substance screening method of this embodiment, the different chemical substances include known chemical substances and unknown chemical substances, and the similarity between the known chemical substance and the unknown chemical substance is determined. . Details of the operation of the similar chemical substance determination unit 48a for determining unknown chemical substances will be explained with reference to the flowchart of FIG. 23.

FIG. 23 is a flowchart showing an example of the operation of determining an unknown chemical substance by the control device 41a according to the present embodiment.
Step S61: The similar chemical substance determining unit 48a compares the quantified data of the unknown chemical substance calculated in step S405 with the quantified data of the known chemical substance registered in the database in step S403. . The similar chemical substance determining unit 48a performs comparison by calculating three indicators. The three indexes here are the IC50 distance, the low concentration difference, and the high concentration difference.

The IC50 distance is the distance between a feature point at a 50% inhibitory concentration for an unknown chemical substance and a feature point at a 50% inhibitory concentration for a known chemical substance in the feature space when the microscope images are captured at the same time. This is calculated between feature points.
The low concentration difference PDL is the sum of the distances between feature points at a concentration lower than 50% inhibitory concentration for an unknown chemical substance, and the distance between feature points at a concentration lower than 50% inhibitory concentration for a known chemical substance. The difference from the sum of the distances is calculated at each imaging time of the microscope image.
The high concentration side difference PDH is the sum of the distances between feature points on the higher concentration side than the 50% inhibitory concentration for an unknown chemical substance, and the features on the higher concentration side of the concentration trajectory than the 50% inhibitory concentration for a known chemical substance. The difference from the sum of the distances between the quantitative points is calculated at each imaging time of the microscope image.

FIG. 24 is a diagram showing an example of a comparison between an unknown chemical substance and a first chemical substance according to the present embodiment. In FIG. 24, the IC50 distance PDI, the low density difference PDL, and the high density difference PDH are shown with respect to the imaging time. The similar chemical substance determination unit 48a calculates the IC50 distance PDI, the low concentration difference PDL, and the high concentration difference PDH.

FIG. 25 is a diagram showing an example of comparison between an unknown chemical substance and a second chemical substance according to the present embodiment. In FIG. 25, the IC50 distance EDI, the low density difference EDL, and the high density difference EDH are shown with respect to the imaging time. The similar chemical substance determining unit 48a calculates the IC50 distance EDI, the low concentration difference EDL, and the high concentration difference EDH.

Returning to FIG. 23, the description of the operation of determining an unknown chemical substance by the control device 41a will be continued. Step S62: The similar chemical substance determining unit 48a selects the weighting of the three indicators calculated in step S61. The similar chemical substance determining unit 48a selects a weight for each of the three indicators. For example, when using only one of the three indicators for determining an unknown chemical substance, the similar chemical substance determination unit 48a sets the weight of this one index to 1, and sets the weight of the other two indexes to 1. Set to zero. Furthermore, when using the average of the three indicators for determining an unknown chemical substance, the similar chemical substance determining unit 48a sets the weights of all three indicators to 1. The similar chemical substance determination unit 48a may arbitrarily select the weights of each of the three indicators.

Step S63: The similar chemical substance determining unit 48a determines the chemical substance most similar to the unknown chemical substance from among the known chemical substances. Here, the similar chemical substance determination unit 48a determines the change in the magnitude of the feature amount at a predetermined concentration at each time point based on the results analyzed by multivariate analysis, and the change in the magnitude of the feature amount at each time point at a predetermined concentration, Different chemistry based on the results of weighting the changes in the size of the feature quantity at each time and the changes in the size of the feature quantity at each time when the concentration is lower than a predetermined concentration. Determine the similarity between substances.

Specifically, the similar chemical substance determination unit 48a calculates the square of the value for each imaging time for each of the three indices calculated in step S61, and calculates the sum of the calculated square values. The similar chemical substance determining unit 48a multiplies the calculated square value by the weight selected in step S62 for each of the three indicators, and calculates the sum. The similar chemical substance determining unit 48a determines that the chemical substance with the smallest calculated sum is the chemical substance that is most similar to the unknown chemical substance. Note that, instead of the sum of square values, the similar chemical substance determination unit 48a may calculate the sum of powers other than squares, or may calculate the sum of absolute values.
In the example shown in FIGS. 24 and 25, the similar chemical substance determination unit 48a distinguishes the first chemical substance (Paclitaxel) from the first chemical substance (Paclitaxel) and the second chemical substance (Etoposide) as an unknown chemical substance. It is determined that the chemical substance is the most similar to the chemical substance.

In this way, the similar chemical substance determining unit 48a determines the similarity between different chemical substances based on changes in feature amounts with respect to changes in a plurality of imaging times after the chemical substance is added to cells at a predetermined concentration. judge. Here, the predetermined concentration includes at least one of a 50% inhibitory concentration, a concentration higher than the 50% inhibitory concentration, and a concentration lower than the 50% inhibitory concentration.

The similar chemical substance determining unit 48a also detects changes in the feature amount with respect to changes in the imaging time at a 50% inhibitory concentration, changes in feature amounts with respect to changes in the imaging time at a concentration higher than the 50% inhibitory concentration, and 50% inhibitory concentration. The similarity between different chemical substances is determined based on the results of weighting the changes in the feature amounts with respect to the changes in the imaging time at a concentration lower than .

In this embodiment, in step S61, the similar chemical substance determination unit 48a determines, as the low concentration difference PDL, the characteristics of the concentration lower than the 50% inhibitory concentration for the unknown chemical substance at each imaging time of the microscope image. Although an example has been described in which the difference between the sum of distances between quantity points and the sum of distances between feature quantity points on the lower concentration side than the 50% inhibitory concentration for a known chemical substance is calculated, the calculation is not limited to the difference. For example, the similar chemical substance determining unit 48a calculates the sum of distances between feature points at a concentration lower than 50% inhibitory concentration for an unknown chemical substance, and the feature value at a concentration lower than 50% inhibitory concentration for a known chemical substance. The ratio of the distance to the sum of distances between points may be calculated, or other known calculations may be performed. Similarly, the similar chemical substance determining unit 48a calculates the sum of the distances between feature points on the higher concentration side than the 50% inhibitory concentration for the unknown chemical substance, and the sum of the distances between the feature points on the higher concentration side than the 50% inhibitory concentration for the known chemical substance. Not limited to the sum and difference of distances between feature points, but also the sum of distances between feature points at a concentration higher than the 50% inhibitory concentration for an unknown chemical substance, and the concentration higher than the 50% inhibitory concentration for a known chemical substance. The ratio between the sum of the distances between the feature amount points on the side may be calculated, or other known calculations may be performed.
Furthermore, in the present embodiment, an example has been described in which the similar chemical substance determining unit 48a selects the weighting of three indicators in step S62, but the process in step S62 may be omitted. If the process in step S62 is omitted, in step S62, the similarity determination unit 48 may calculate, for example, the sum of square values for each of the three indicators. Furthermore, in the present embodiment, a case has been described in which the similar chemical substance determining unit 48a determines the chemical substance with the smallest calculated sum as the chemical substance most similar to the unknown chemical substance in step S63. It is not limited to this. If the calculated sum is larger than a predetermined threshold, the similar chemical substance determination unit 48a may determine that the chemical substance to be determined is not similar to the unknown chemical substance.

In this embodiment, in step S402 and step S405, the time trajectory quantification unit 47a quantifies the feature points in the feature space by calculating the sum of the distances between the feature points on the higher concentration side than the 50% inhibitory concentration. , and the sum of distances between feature points on the lower concentration side than the 50% inhibition concentration have been described, but the present invention is not limited thereto. In step S402, the time trajectory quantification unit 47a calculates the sum of distances between feature points on the higher concentration side than the 50% inhibitory concentration, and calculates the sum of distances between the feature points on the higher concentration side than the 50% inhibitory concentration, as quantification of the feature points in the feature space. At least one process of calculating the sum of distances between feature points on the density side may be performed. In step S402 and step S405, the time trajectory quantification unit 47a calculates the sum of distances between feature points on the higher concentration side than the 50% inhibition concentration as quantification of the feature points in the feature space, and When performing at least one process of calculating the sum of distances between feature points on the lower density side than the density, in accordance with the process of quantifying feature points in the feature space performed by the time trajectory quantification unit 47a, In step S406, the similar chemical substance determining unit 48a determines whether the unknown chemical substance is a known chemical substance based on at least one of the three indicators: IC50 distance, low concentration difference, and high concentration difference. A determination is made as to whether it is most similar to. Furthermore, the 50% inhibition concentration may not be used as a reference in calculating the distance between feature points. For example, the time trajectory quantification unit 47a calculates, with a predetermined density as a reference, the sum of distances between feature points on the higher density side than the predetermined density, and/or the sum of distances between feature points on the lower density side. may be calculated and the feature points may be quantified. Further, although an example has been described in which the time trajectory quantification unit 47a calculates the distance (IC50 distance) between feature points at 50% inhibitory concentration in step S402 and step S405, the present invention is not limited to this. In step S402 and step S405, the time trajectory quantification unit 47a may calculate the distance between feature points for each imaging time of the microscope image for concentrations other than the 50% inhibitory concentration.

≪Summary≫
As explained above, in the chemical substance screening method provided by the control device 41a of the present embodiment, microscopic images are captured at a plurality of imaging times after the chemical substance is added to the cells, and at a predetermined concentration. The similarity between different chemical substances is determined based on changes in feature amounts with respect to changes in imaging time.
With this configuration, in the chemical substance screening method provided by the control device 41a of the present embodiment, morphological changes caused by chemical substances in cells can be determined by inexpensive and quick analysis without fixing or staining cells. This allows us to determine the similarity between chemicals that cause morphological changes in cells at a given concentration using an inexpensive and rapid analysis without fixing or staining the cells. I can do it.

Further, in the chemical substance screening method provided by the control device 41a of the present embodiment, the predetermined concentration is at least one of a 50% inhibitory concentration, a concentration higher than the 50% inhibitory concentration, and a concentration lower than the 50% inhibitory concentration. Contains one concentration.
With this configuration, in the chemical substance screening method provided by the control device 41a of the present embodiment, morphological changes caused by chemical substances in cells over time can be evaluated by inexpensive and rapid analysis without fixing or staining cells. % inhibitory concentration, a concentration higher than the 50% inhibitory concentration, and a concentration lower than the 50% inhibitory concentration. Chemical substances that cause a morphological change when the concentration of the added chemical substance is at least one of a 50% inhibitory concentration, a concentration higher than the 50% inhibitory concentration, and a concentration lower than the 50% inhibitory concentration. Similarity can be determined.

In addition, the chemical substance screening method provided by the control device 41a of the present embodiment includes changes in the feature amount with respect to changes in the imaging time at a 50% inhibitory concentration, and characteristics with respect to changes in the imaging time at a concentration higher than the 50% inhibitory concentration. The similarity between different chemical substances is determined based on the results of weighting the change in amount and the change in feature amount with respect to change in imaging time at a concentration lower than the 50% inhibitory concentration.
With this configuration, in the chemical substance screening method provided by the control device 41a of the present embodiment, a change in feature amount with respect to a change in imaging time at a 50% inhibitory concentration, and a change in a feature amount with respect to a change in imaging time at a 50% inhibitory concentration, The similarity between chemical substances is calculated by weighting the change in the feature amount in response to a change in imaging time at a concentration higher than the 50% inhibition concentration and the change in the feature amount in response to a change in imaging time at a concentration lower than the 50% inhibition concentration. Since it is possible to determine the gender of chemical substances, the accuracy of screening for chemical substances can be improved compared to when weighting is not performed.

Since the chemical substance screening method of the present invention is a non-destructive system, it is suitable for use in time-lapse analysis, and the sample used for screening itself can be subjected to another evaluation system, and the screening results can be transferred to another evaluation system. It can be evaluated from

Note that a program for executing each process of the culture observation device 11 and the culture observation device 11a in the embodiment of the present invention is recorded on a computer-readable recording medium, and the program recorded on the recording medium is transferred to a computer system. The various processes described above may be performed by loading and executing the program.

Note that the "computer system" here may include hardware such as an OS and peripheral devices. Furthermore, the term "computer system" includes the homepage providing environment (or display environment) if a WWW system is used. Furthermore, "computer-readable recording media" refers to flexible disks, magneto-optical disks, ROMs, writable non-volatile memories such as flash memories, portable media such as CD-ROMs, hard disks built into computer systems, etc. storage device.

Furthermore, "computer-readable recording medium" refers to volatile memory (for example, DRAM (Dynamic It also includes those that retain programs for a certain period of time, such as Random Access Memory). Further, the program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in a transmission medium. Here, the "transmission medium" that transmits the program refers to a medium that has a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. Moreover, the above program may be for realizing a part of the above-mentioned functions. Furthermore, it may be a so-called difference file (difference program) that can realize the above-described functions in combination with a program already recorded in the computer system.

Although the embodiment of the present invention has been described above in detail with reference to the drawings, the specific configuration is not limited to this embodiment, and includes designs within the scope of the gist of the present invention.

11, 11a... Culture observation device, 41, 41a... Control device, 44, 44a... Feature value calculation unit, 45, 45a... Coordinate calculation unit, 46... Concentration trajectory calculation unit, 47... Change area determination unit, 48... Similarity Determination unit, 46a... Time trajectory calculation unit, 47a... Time trajectory quantification unit, 48a... Similar chemical substance determination unit, 50... Display unit

Claims (23)

Obtaining a plurality of feature amounts regarding the morphology of the cells included in microscopic images of cells in a plurality of cell groups to which different chemical substances have been added at different concentrations;
The distance between feature points, which are points specified by a set of values indicating each of the plurality of feature quantities, in a feature space, which is a space in which coordinate axes are defined for each type of plurality of feature quantities related to the morphology of the cell. , comparing the different chemical substances in a morphological change region that is a concentration range of the chemical substance in which the magnitude of change in the feature value point with respect to a change in the concentration of the chemical substance added to the cell is a predetermined value or more; determining the similarity between the different chemical substances by ;
A screening method for chemical substances having further comprising determining the morphological change region by excluding a control region and a dead cell region from a range of concentrations of a chemical substance added to the cells;
The control region is a concentration range in which, when the concentration of the chemical substance added to the cell is increased from the lowest concentration, the change in the feature amount with respect to the concentration of the chemical substance is smaller than the predetermined value. ,
The dead cell region is a region where when the concentration of a chemical substance added to the cell is increased, the cell dies and the change in the feature amount with respect to the concentration of the chemical substance becomes smaller than the predetermined value. concentration range
A method for screening chemical substances according to claim 1. determining the control region based on the distance of the feature point from the center of the control region in the feature space;
and determining the dead cell area based on the distance of the feature point from the center of the dead cell area in the feature space.
The method for screening chemical substances according to claim 2. The plurality of feature amounts regarding the morphology of the cells include a feature amount regarding the morphology of each of the cells, a feature amount regarding the morphology of the cell group as a group, and a temporal change in the feature amount regarding the morphology of the cell group as a group. including
A method for screening chemical substances according to claim 1. 2. The method of screening a chemical substance according to claim 1 , wherein the morphological change region includes a concentration range from when the cell's morphology changes until the cell dies. 6. The chemical substance screening method according to claim 1 , further comprising determining the morphological change region based on a change in the feature amount with respect to a change in the concentration. 7. The chemical substance screening method according to claim 6 , wherein determining the morphological change region includes determining the morphological change region by comparing the amount of change in the feature amount with a threshold value. The microscopic images are taken at a plurality of imaging times after the chemical substance is added to the cells,
The chemical substance according to any one of claims 1 to 7 , wherein the similarity between the different chemical substances is determined based on a change in the feature amount with respect to a change in the imaging time at a predetermined concentration. screening method. 9. The method for screening a chemical substance according to claim 8 , wherein the predetermined concentration includes at least one of a 50% inhibitory concentration, a concentration higher than the 50% inhibitory concentration, and a concentration lower than the 50% inhibitory concentration. a change in the feature amount with respect to a change in the imaging time at the 50% inhibitory concentration; a change in the feature amount with respect to a change in the imaging time at a concentration higher than the 50% inhibitory concentration; and a change in the feature amount with respect to a change in the imaging time at a concentration higher than the 50% inhibitory concentration. 10. The chemical substance screening method according to claim 9 , wherein the similarity is determined based on a result of weighting a change in the feature quantity with respect to a change in the imaging time in concentration. The different chemical substances include known chemical substances and unknown chemical substances,
The chemical substance screening method according to any one of claims 1 to 10 , wherein the similarity between the known chemical substance and the unknown chemical substance is determined. The method for screening chemical substances according to any one of claims 1 to 11 , wherein the similarity is determined based on a result of analyzing the plurality of feature amounts by multivariate analysis. to the computer,
A program for executing steps based on the chemical substance screening method according to any one of claims 1 to 12 . a feature amount calculation unit that calculates a plurality of feature amounts related to the morphology of the cells included in microscopic images of cells in a plurality of cell groups to which different chemical substances are added at different concentrations;
The distance between feature points, which are points specified by a set of values indicating each of the plurality of feature quantities, in a feature space, which is a space in which coordinate axes are defined for each type of plurality of feature quantities related to the morphology of the cell. , comparing the different chemical substances in a morphological change region that is a concentration range of the chemical substance in which the magnitude of change in the feature value point with respect to a change in the concentration of the chemical substance added to the cell is a predetermined value or more; a similarity determination unit that determines the similarity between the different chemical substances;
A control device comprising: 15. The control device according to claim 14 , wherein the morphology change region includes a concentration range from when the morphology of the cells changes until the cells die. The control device according to claim 14 or claim 15 , further comprising a density determination unit that determines the shape change region based on a change in the feature amount with respect to a change in the density. 17. The control device according to claim 16 , wherein the density determination unit determines the shape change region by comparing the amount of change in the feature amount with a threshold value. The microscopic images are taken at a plurality of imaging times after the chemical substance is added to the cells,
Any one of claims 14 to 17 , wherein the similarity determination unit determines the similarity between the different chemical substances based on a change in the feature amount with respect to a change in the imaging time at a predetermined concentration. The control device according to item 1. The control device according to claim 18 , wherein the predetermined concentration includes at least one of a 50% inhibitory concentration, a concentration higher than the 50% inhibitory concentration, and a concentration lower than the 50% inhibitory concentration. The similarity determining unit determines a change in the feature amount in response to a change in the imaging time at the 50% inhibition concentration, a change in the feature amount in response to a change in the imaging time at a concentration higher than the 50% inhibition concentration, and a change in the feature amount in response to a change in the imaging time at a concentration higher than the 50% inhibition concentration. 20. The control device according to claim 19 , wherein the similarity is determined based on a result of weighting a change in the feature quantity with respect to a change in the imaging time at a concentration lower than a 50% inhibitory concentration. The different chemical substances include known chemical substances and unknown chemical substances,
The control device according to any one of claims 14 to 20 , wherein the similarity between the known chemical substance and the unknown chemical substance is determined. The control device according to any one of claims 14 to 21 , wherein the similarity determination unit determines the similarity based on a result of analyzing the plurality of feature amounts by multivariate analysis. A culture observation device comprising the control device according to any one of claims 14 to 22 .

Images