A method, system and storage medium for cryo-em data collection

Abstract

This application provides a method, system, and storage medium for cryo-electron microscopy data collection. The method includes: acquiring a sample map on a cryo-electron microscope grid; the sample map includes at least one grid, each grid including at least one circular hole; determining a first circular hole in the sample map that meets preset conditions based on the matching of at least one preset mask with the sample map; identifying all target particles contained in the first circular hole using a preset recognition model; and determining the image data acquisition position based on the positional characteristics of each target particle in the circular hole. This application can more accurately determine the position of the circular hole and the position of the target particles within the circular hole. By determining the position of the target particles within the circular hole, the optimal imaging position can be determined. The entire process requires no manual intervention, saving a significant amount of time. Furthermore, the determined imaging positions of the target particles are more precise, and a larger number of usable particles are obtained, which can greatly improve the quality and quantity of data collection.

Description

Technical Field

[0001] This application relates to the field of cryo-electron microscopy, and in particular to a cryo-electron microscopy data collection method, system, and storage medium. Background Technology

[0002] Cryo-electron microscopy is primarily used in structural biology, providing high-resolution structures for proteins, macromolecular complexes, and viruses, elucidating the structure-function relationship at the atomic level. Data is collected using cryo-electron microscopy, and then single-particle structure analysis is used to obtain the three-dimensional structure of the sample. The results of structure analysis depend on the number and quality of particles in the collected data. However, some macromolecular complexes and viruses exhibit uneven distribution within the pores during sample preparation; for example, many empty pores may exist, or the particles may be unevenly distributed and dispersed within the pores.

[0003] Because cryo-electron microscopy can only acquire one image per aperture, the data collection location (i.e., the imaging location) needs to be determined in advance to ensure that the acquired image contains as many particles as possible. Traditionally, users determine the data collection location primarily through automatic aperture selection and manual aperture selection methods.

[0004] 1. Automatic hole selection method.

[0005] The center of the circular hole was used as the data collection location. The automatic hole selection software SerialEM was employed, based on the positional relationships between holes. For example, (a) five points were first selected, with the center of these five manually added holes marked as X, and the positions of the other holes marked as o; (b) then the positions of all other points were calculated using these five points. (c) Automatic hole selection based on image recognition was used, with the hole center as the data collection location. However, when the sample particles are unevenly distributed, the automatic hole selection method may result in the absence of target sample particles in the imaging area, leading to data collection failure.

[0006] 2. Manual hole selection method.

[0007] The location of particles within the circular holes is manually determined, and then shooting points are determined near the particles. The positions of the crosshairs are manually added. However, the manual hole selection method has two main problems: First, it requires manually adding points one by one, which is time-consuming. Second, when shooting in groups of circular holes, it is difficult to achieve the optimal state, resulting in slower data collection speed. Moreover, the distance between some holes and the center hole is too large, reducing the quality of data collection. Summary of the Invention

[0008] This application is made in view of at least one of the above-mentioned technical problems existing in the prior art. According to one aspect of this application, a cryo-electron microscopy data acquisition method is provided, the method comprising:

[0009] Collect a sample map on a cryo-electron microscope grid; the sample map includes at least one grid, and each grid includes at least one circular hole;

[0010] Based on the matching between at least one preset mask and the sample map, the first circular hole in the sample map that meets the preset conditions is determined;

[0011] Using a preset recognition model, all target particles contained in the first circular hole are identified;

[0012] The location for acquiring image data is determined based on the positional characteristics of each target particle within the circular hole.

[0013] In some embodiments, determining a first circular hole in the sample map that meets preset conditions based on the matching between at least one preset mask and the sample map includes:

[0014] Each circular hole in the sample map is sequentially matched with at least one preset mask;

[0015] When a match is successful, each of the circular holes is the first circular hole that meets the preset conditions.

[0016] In some embodiments, the method further includes:

[0017] The preset mask includes at least one of the following: a complete ring, a 1 / 2 ring, and a 1 / 4 ring having a preset outer circle radius and a preset inner circle radius.

[0018] In some embodiments, the method further includes:

[0019] The sample data is labeled and trained to construct the preset recognition model;

[0020] The preset recognition model is tested to determine whether it meets the preset requirements.

[0021] In some embodiments, determining the image data acquisition location based on the positional characteristics of each target particle within the circular hole includes:

[0022] Using each target particle as the center of a single acquisition area, determine the number of target particles contained in the single acquisition area corresponding to each target particle.

[0023] The area containing the largest number of target particles in a single sampling session is selected as the sampling location.

[0024] In some embodiments, determining the image data acquisition location based on the positional characteristics of each target particle within the circular hole includes:

[0025] Construct the bounding rectangle of all target particles within the first circular hole;

[0026] Define a single-sampling area within the circumscribed rectangle;

[0027] Within the bounding rectangle, the single-collection area is traversed and moved with a preset step size;

[0028] Determine the number of target particles in the single acquisition area during each movement;

[0029] The area containing the largest number of target particles in a single sampling session is selected as the sampling location.

[0030] In some embodiments, when at least two single-collection regions contain the maximum number of target particles, the method further includes:

[0031] Determine the distance between the center of each single acquisition area and the center of the first circular hole;

[0032] The sampling location is determined by identifying the area closest to the center of the circular hole in a single sampling operation.

[0033] In some embodiments, the sample map includes at least two first circular holes, and the method further includes:

[0034] The at least two first circular holes are grouped according to a preset rule;

[0035] Data is collected for each of the first circular holes in each group according to the stated collection position.

[0036] Another embodiment of this application provides a cryo-electron microscopy data collection system, the system comprising:

[0037] The system includes a memory and a processor, wherein the memory stores a computer program that is executed by the processor, and the computer program, when executed by the processor, causes the processor to perform the cryo-electron microscopy data collection method as described above.

[0038] In another aspect, this application provides a storage medium storing a computer program that, when run by a processor, causes the processor to execute the cryo-electron microscopy data collection method as described above.

[0039] The cryo-electron microscopy data collection method of this application determines the first circular hole in the sample map that meets the preset conditions by matching the preset mask with the sample map. Then, a preset recognition model is used to identify all target particles contained in the first circular hole. Based on the positional characteristics of each target particle in the circular hole, the image data acquisition position is determined. This can more accurately determine the position of the circular hole and the position of the target particles in the circular hole. By determining the position of the target particles in the circular hole, the optimal shooting position is determined. The whole process does not require manual intervention, saving a lot of time. At the same time, the shooting position of the determined target particles is more accurate, and more particles are available, which can greatly improve the quality and quantity of data collection. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 A schematic flowchart illustrating a cryo-electron microscopy data collection method according to an embodiment of this application is shown;

[0042] Figure 2 A schematic diagram of a map according to an embodiment of this application is shown;

[0043] Figures 3(1) to 3(4) This diagram illustrates the results of anomaly identification of circular holes using traditional techniques.

[0044] Figures 4(1) to 4(7) A schematic diagram of a preset mask according to an embodiment of this application is shown;

[0045] Figure 5 A schematic flowchart illustrating step S102 according to an embodiment of this application is shown;

[0046] Figures 6(1) to 6(4) Show respectively Figures 3(1) to 3(4) The diagram shows the situation where the circular hole matches the preset mask.

[0047] Figure 7 A schematic flowchart illustrating the construction of the preset recognition model according to an embodiment of this application is shown;

[0048] Figure 8 A schematic flowchart illustrating step S104 according to one embodiment of this application is shown;

[0049] Figure 9 A schematic flowchart illustrating step S104 according to another embodiment of this application is shown;

[0050] Figure 10 A schematic diagram showing the circumscribed rectangle of all target particles within a circular hole according to an embodiment of this application;

[0051] Figure 11 A schematic diagram showing the result of determining the shooting position using an algorithm that traverses the bounding rectangle of all target particles according to an embodiment of this application;

[0052] Figure 12 A schematic flowchart illustrating the grouping of circular holes according to an embodiment of this application is shown;

[0053] Figure 13 A schematic diagram showing the distance between any circular hole and the center circular hole according to an embodiment of this application;

[0054] Figure 14 This diagram illustrates how all the circular holes are divided into a plurality of square grids according to an embodiment of the present application.

[0055] Figure 15 The diagram shown is a grouping of 3*3 circular holes according to an embodiment of this application;

[0056] Figure 16 A schematic flowchart illustrating a cryo-electron microscopy data collection method according to yet another embodiment of this application is shown;

[0057] Figure 17 A schematic block diagram of a cryo-electron microscopy data collection system according to an embodiment of this application is shown. Detailed Implementation

[0058] To enable those skilled in the art to better understand the technical solutions of the embodiments of this application, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0059] Based on at least one of the aforementioned technical problems, this application provides a cryo-electron microscopy data collection method. The method includes: acquiring a sample map on a cryo-electron microscope grid; the sample map includes at least one grid, each grid including at least one circular hole; determining a first circular hole in the sample map that meets preset conditions based on the matching between at least one preset mask and the sample map; identifying all target particles contained in the first circular hole using a preset recognition model; and determining the image data acquisition position based on the positional characteristics of each target particle in the circular hole. This cryo-electron microscopy data collection method, according to an embodiment of this application, determines a first circular hole in the sample map that meets preset conditions by matching between a preset mask and the sample map, then identifies all target particles contained in the first circular hole using a preset recognition model, and finally determines the image data acquisition position based on the positional characteristics of each target particle in the circular hole. This allows for more accurate determination of the position of the circular hole and the position of the target particles within it. The optimal imaging position is determined based on the position of the target particles within the circular hole. The entire process requires no manual intervention, saving significant time. Furthermore, the determined imaging positions of the target particles are more precise, and the number of usable particles is greater, significantly improving the quality and quantity of data collection.

[0060] Figure 1 A schematic flowchart illustrating a cryo-electron microscopy data collection method according to an embodiment of this application is shown; Figure 1 As shown, the cryo-electron microscopy data collection method 100 according to an embodiment of this application may include the following steps S101, S102, S103 and S104:

[0061] In step S101, a sample map is acquired on the cryo-electron microscope grid; the sample map includes at least one grid, and each grid includes at least one circular hole.

[0062] In this embodiment, the grid carrying the frozen sample is first loaded into a cryo-electron microscope. A sample map is then captured, which may include a large map and a medium map. The large map refers to the overall image containing the grid carrying the biological sample, while the medium map refers to an image containing a single grid cell on the grid.

[0063] In this embodiment, a medium-sized map is primarily used to determine the location of the circular hole more accurately. The medium-sized map can be exported as a JPG image. Figure 2 The image shown is a schematic diagram of the central map.

[0064] In step S102, based on the matching between at least one preset mask and the sample map, a first circular hole in the sample map that meets the preset conditions is determined.

[0065] In traditional techniques, the coordinates of circular holes on a carrier mesh can be identified using Hough circle recognition and template matching algorithms. For small protein molecules evenly distributed within the holes, the center of the hole can be used as the imaging position. However, for large molecular complexes such as viruses, the sample preparation process involves non-uniformity and aggregation on the carrier mesh, resulting in contaminated areas in the captured images. Therefore, relying solely on Hough circle recognition and template matching algorithms can lead to a large number of holes containing target particles not being selected, or the selected positions being incorrect. Figures 3(1) to 3(4) The figure shows a schematic diagram of the abnormal identification results of circular holes in the traditional technology. In Figures 3(1), 3(2) and 3(4), the circular holes are dirty (e.g., there is ice contamination), which causes the circular holes to not match the template. In Figure 3(2), the circular hole cannot be identified because the image of the circular hole is dark.

[0066] In one embodiment of this application, such as Figure 5 As shown, step S102, which determines the first circular hole in the sample map that meets preset conditions based on the matching between at least one preset mask and the sample map, includes steps A1 and A2:

[0067] In step A1, each circular hole in the sample map is sequentially matched with the at least one preset mask;

[0068] In step A2, when a match is successful, each circular hole is the first circular hole that meets the preset conditions.

[0069] To more accurately identify the location of the circular hole, this application embodiment introduces multiple preset masks. The circular holes in the map are matched against these preset masks. During the matching process, if no matching hole is found, the next preset mask is used for matching until a matching hole is found. At this point, the circular hole is considered to meet the criteria.

[0070] The preset mask includes at least one of the following: a complete annulus, a half-annulus, and a quarter-annulus having a preset outer radius and a preset inner radius. For example... Figures 4(1) to 4(7) As shown, Figure 4(1) shows the entire ring, while Figures 4(2) and 4(3) show half of the ring. Figures 4(4) to 4(7) It is a 1 / 4 ring.

[0071] Combined with Figure 4 and Figure 5The mask's function is to use only the important parts of the circular hole in the similarity calculation. For example, in the process of using the mask to identify the circular hole, whether there are virus particles or ice contamination inside the circular hole is not a concern. In the first mask shown in Figure 4(1), circles are drawn with the center of the circular hole as the center o (the radius r of the inner hole), and the radius r+d and radius rd as the radii. Since the selected template is a square tangent to the top, bottom, left, and right of the circular hole, and the circular hole is circular, the square is not large enough to accommodate all the circles with radius r+d. Therefore, at the top, bottom, left, and right positions, the circles with radius r+d are missing at the edges of the square. Figures 4(2) to 4(7) Similarly, for ease of description, we will continue to refer to it as a ring. The pixel value of the ring between the two circles is assigned 255 (255 represents white in a grayscale image), and the pixels in other areas are assigned 0 (0 represents black in a grayscale image). During the template matching calculation, only the white area, i.e., the ring part in the hole, participates in the similarity calculation; the content of other areas is masked and does not affect the hole matching result. However, the boundary of the hole in the map of the frozen sample is still complex, and there is a high probability that there are large molecular complexes or ice contamination at the hole boundary, which means that even after using the ring as a mask, a satisfactory result cannot be obtained. We continue to use a semi-circular ring mask, a 1 / 4 circular ring mask (such as... Figures 4(2) to 4(7) (As shown) Continue attempting to match. By using the different masks described above for template matching, in most cases, a satisfactory hole matching result can be obtained.

[0072] like Figures 6(1) to 6(4) As shown, they are respectively Figures 3(1) to 3(4) This is a schematic diagram showing the matching of the circular hole with the preset mask. (From...) Figures 6(1) to 6(4) It can be seen that even when the holes are dirty (e.g., there is ice contamination) or the image is dark, it can still match very well.

[0073] In step S103, a preset recognition model is used to identify all target particles contained in the first circular hole.

[0074] In one embodiment of this application, such as Figure 7 As shown, the method further includes steps B1 and B2:

[0075] In step B1, the sample data is labeled and trained to construct the preset recognition model;

[0076] In step B2, the preset recognition model is tested to determine whether the preset recognition model meets the preset requirements.

[0077] In this embodiment, after identifying the circular hole in step S102, it is also necessary to accurately identify the position of the target particle within the hole. However, the target particle is usually small, and traditional template matching algorithms will have significant errors. To achieve better target particle identification and localization, this application employs a deep learning model, and the YOLO v8 algorithm is used as an example for description.

[0078] First, the sample data needs to be labeled. To accurately identify target particle information, a large number of particles need to be labeled. For example, an online labeling tool (Roboflow) can be used to manually circle the target particles with squares on the image. After all target particles are labeled, the dataset is split into a training set, a validation set, and a test set according to a certain ratio. All the split data is downloaded locally to prepare for subsequent training.

[0079] Secondly, the preset recognition model is trained. A mainstream deep learning model (YOLOv8 as an example) can be used for training. The neural network model parameters are set, the training set is input into the neural network model, and the model is trained based on YOLOv8. The model is saved every n rounds during training. The model with the smallest loss function value is selected as the optimal model using the test set.

[0080] Finally, the trained preset recognition model is used to predict the results. The model trained in step 2 above is used to detect large molecular particles inside the circular hole, predict the particle positions, and output the prediction results. The results are saved as a text file or stored in a database.

[0081] In step S104, the image data acquisition location is determined based on the positional characteristics of each target particle in the circular hole.

[0082] In one embodiment of this application, such as Figure 8 As shown, step S104, which determines the image data acquisition location based on the positional characteristics of each target particle in the circular hole, includes steps C1 and C2:

[0083] In step C1, taking each target particle as the center of a single acquisition area, the number of target particles contained in the single acquisition area corresponding to each target particle is determined.

[0084] In step C2, the area containing the largest number of target particles in a single sampling is selected as the sampling location.

[0085] The following example illustrates how to determine the shooting location using a traversal algorithm centered on the target particle.

[0086] At different magnifications, the size of the area that can be captured by cryo-electron microscopy varies. Taking a cryo-electron microscope with a magnification of 81,000x as an example, the captured area is approximately 0.62μm × 0.44μm. Therefore, the main goal of this step is to select a region of length μm × width μm from each well selected above, matching the magnification of the data collection, so that this region contains the maximum number of target particles. The steps are as follows:

[0087] (1) Sort all target particles identified within the circular hole according to their coordinates (x, y). The sorting result is ((x1, y1), (x2, y2), ... (x...). n ,y n )).

[0088] (2) Iterate through each target particle in turn, taking the coordinates of that target particle as the center, and...

[0089] A rectangle with four vertices.

[0090] (3) Iterate through all the particles in the hole in turn, and calculate whether each particle is located within the rectangle mentioned above, so as to count the total number of target particles located within the rectangle. n .

[0091] (4) Count all particles according to the number of particles in their corresponding regions. n Sort the data. Select count. n The coordinates of the largest particle are used as the coordinates to be photographed.

[0092] (5) If two or more particles have the same count, then select the coordinates of the particle closest to the center of the hole as the coordinates to be photographed, based on the distance between the particle and the center of the hole. This is because the ice layer inside the hole is generally thinner than at the edge of the hole, making it easier to obtain a good photo.

[0093] In this embodiment of the application, when the target particles inside the circular hole are relatively sparse, the traversal algorithm centered on the target particles will be faster.

[0094] In another embodiment of this application, such as Figure 9 As shown, step S104, which determines the image data acquisition location based on the positional characteristics of each target particle in the circular hole, includes steps D1, D2, D3, D4, and D5:

[0095] In step D1, construct the circumscribed rectangle of all target particles within the first circular hole;

[0096] In step D2, a single acquisition area is defined within the circumscribed rectangle;

[0097] In step D3, the single-collection area is traversed and moved within the bounding rectangle with a preset step size;

[0098] In step D4, the number of target particles in the single acquisition area is determined during each movement;

[0099] In step D5, the area containing the largest number of target particles in a single sampling is selected as the sampling location.

[0100] The following example illustrates how to determine the shooting location using an algorithm that traverses the bounding rectangle of all target particles. The steps are as follows:

[0101] (1) Select the bounding rectangle of all target particles, such as Figure 10 As shown. The coordinates of the circumscribed rectangle are ((recx)). start ,recy start ),(recx end ,recy start ),(recx end ,recy end ),(recx start ,recy end ))

[0102] (2) Loop through all regions within the bounding rectangle, in the x-direction, recx start Starting from s, with a step size of s x , with recx end -length is the endpoint, and the nth step is recx. start +n×s x In the y-direction, with recy start Starting from s, with a step size of s y , with recy end -width is the endpoint, and the m-th step is recy start +m×s y Then with (recx) start +n×s x ,recy start +m×s y Let '(')' be the top-left vertex, with length and width respectively. Generate a rectangle, denoted as Rec. mn .

[0103] (3) Iterate through all the particles in the hole, calculate whether the particle is located within the rectangle, and count the total number of target particles located within the rectangle. mn .

[0104] (4) Count all rectangles according to the number of particles in their corresponding areas. mn Sort the data. Select count. mn Find the largest region. Determine the center point of this region and use it as the shooting point.

[0105] (5) If two or more regions have the same count, then based on the distance between the center of each region and the center of the hole, select the coordinates of the center point of the region closest to the center of the hole as the coordinates to be photographed. This is because the ice layer in the middle region of the hole is generally thinner than that in the edge region of the hole, making it easier to obtain a good photo.

[0106] like Figure 11 The image shows the result of determining the shooting position using an algorithm that traverses the bounding rectangle of all target particles. This algorithm allows for more precise location of the area with the highest number of target particles.

[0107] In one embodiment of this application, such as Figure 12 As shown, the sample map contains at least two first circular holes, and the method further includes steps E1 and E2:

[0108] In step E1, the at least two first circular holes are grouped according to a preset rule;

[0109] In step E2, data is collected for each first circular hole in each group according to the collection position.

[0110] Through steps S101 to S104, the positional information of the macromolecular complex particles within the circular holes in the central image and the determined photographic positions within the holes can be obtained. To ensure the quality and efficiency of data collection, multiple circular holes can be grouped, and then photographed group by group. For example, the group size can be set to 5*5, and the distance from the center hole to the edge within the group can be within 8 μm. While meeting the above conditions, in order to improve the photographing speed, there should be as many holes as possible in each group on average.

[0111] It is worth noting that, given the non-uniform distribution of frozen samples of macromolecular complexes, there are a large number of empty circular holes. Removing empty circular holes and directly grouping them would result in fewer average photographable holes within a single group, thus slowing down the imaging speed. Therefore, it is not necessary to remove empty circular holes during the grouping stage.

[0112] Furthermore, within the same group, the distance between any circular hole and the central circular hole should not be too far; otherwise, the cryo-electron microscopy imaging results will be affected. Figure 13As shown, the distance between two circular holes is generally 2.5 μm, the distance d1 between holes 20 and 29 is approximately 12.5 μm, and the distance d2 between holes 22 and 29 is approximately 10 μm. The distances between holes 20 and 29, and between holes 22 and 29, are too large. Therefore, using any one of these three holes as the center point, the three holes can only be divided into two groups. To solve this problem, it is necessary to use all the identified circular holes for reasonable grouping. For example... Figure 14 The diagram shown illustrates how all the circular holes are divided into several square grids.

[0113] like Figure 15 The diagram shown illustrates the process of grouping holes into 3x3 grids. The steps are as follows:

[0114] (1) Using the grid grouping method, the three circular holes can be grouped into the same group, with the central circular hole as the center point, recorded as P1. The distance from P1 to circular hole 20 is approximately 7.07 μm, the distance from P1 to circular hole 22 is approximately 5.59 μm, and the distance from P1 to circular hole 29 is approximately 5.59 μm. The above method introduces an empty circular hole as the center circular hole of the group, so that the three circular holes are in the same group, and the distance from each circular hole to the center circular hole is within the required range.

[0115] (2) If there is a target particle in the hole where P1 is located, then the coordinates of the center hole are the center of the shooting area determined in the above steps. If there is no target particle in P1, then the center position of the hole where P1 is located is determined as the position of P1, and P1 is marked as empty.

[0116] (3) Set the threshold for the minimum number of shooting positions within the group as t.

[0117] (4) Traverse all the holes in the group and determine the number of locations in the group that can be used to collect data. If the number of locations in the group that can be photographed is less than t, then delete the group; otherwise, keep the group.

[0118] (5) Clean all empty holes in the group except for the central hole.

[0119] After grouping, the coordinates of the final selected holes are converted into a coordinate file that can be recognized by the automatic hole selection software (SerialEM). When the center point of each group is marked as empty, it means that there are no particles at the current position, and it is only used as the target position for stage movement, without taking pictures to collect data.

[0120] The cryo-electron microscopy data collection method of this application determines the first circular hole in the sample map that meets the preset conditions by matching the preset mask with the sample map. Then, a preset recognition model is used to identify all target particles contained in the first circular hole. Based on the positional characteristics of each target particle in the circular hole, the image data acquisition position is determined. This can more accurately determine the position of the circular hole and the position of the target particles in the circular hole. By determining the position of the target particles in the circular hole, the optimal shooting position is determined. The whole process does not require manual intervention, saving a lot of time. At the same time, the shooting position of the determined target particles is more accurate, and more particles are available, which can greatly improve the quality and quantity of data collection.

[0121] like Figure 16 The diagram shown is a schematic flowchart of a cryo-electron microscopy data collection method according to another embodiment of this application. The cryo-electron microscopy data collection method 1600 according to an embodiment of this application may include the following steps: S1601, S1602, S1603, S1604, S1605, S1606, S1607, S1608, S1609, S1610, S1611, S1612, S1613, S1614, and S1615.

[0122] In step S1601, a mid-map of the sample on the cryo-electron microscope grid is acquired.

[0123] In step S1602, the map is exported as a JPG file.

[0124] In step S1603, the center map is matched with the preset mask to obtain the best matching circular hole.

[0125] Due to the non-uniformity and aggregation of samples such as large molecular complexes and viruses during sample preparation, the areas captured in the mid-map are often quite dirty. Therefore, using only the Hough circle recognition algorithm and template matching algorithm will result in many holes containing target particles not being selected. To more accurately identify the location of the holes, this application introduces multiple masks, matching the holes in the mid-map with the masks. During the matching process, if no matching hole is found, the next mask is used until a matching hole is found. At this point, the hole is considered to meet the criteria. The preset mask includes at least one of the following: a complete ring, a 1 / 2 ring, and a 1 / 4 ring with preset outer and inner radii.

[0126] In step S1604, the identified circular holes are extracted as individual images.

[0127] In step S1605, an image of any one of the circular holes is obtained.

[0128] In step S1606, the size and location of the target particles inside the pore are marked.

[0129] In step S1607, all labeled holes are split into training set, test set and validation set.

[0130] In step S1608, a preset recognition model for the target particles is trained based on a deep learning algorithm.

[0131] In step S1609, the optimal preset recognition model is determined through multiple iterations.

[0132] Steps S1605 to S1609 above constitute the process of training a preset recognition model. This preset recognition model can be pre-trained before this method so that it can be used directly when determining the shooting position.

[0133] In step S1610, the shooting position is determined by the target particle position identified by the preset recognition model.

[0134] In step S1611, the holes are grouped based on their positions.

[0135] In step S1612, the center point of each group is determined, and the center point is used as the first point of each group to mark whether the circular hole is empty.

[0136] In step S1613, other empty holes in the group are filtered out.

[0137] In this embodiment, all circular holes at the center point, regardless of whether they are empty, should be retained to ensure accurate location determination. However, the circular holes at the center point will be marked as empty to avoid data acquisition. Empty circular holes not at the center point should be deleted without processing to improve shooting efficiency.

[0138] In step S1614, the data is exported as a navigation file that SerialEM can recognize.

[0139] In step S1615, image data of the target particles is acquired.

[0140] The purpose of this invention is to provide a method for accurately selecting data collection locations for cryo-electron microscopy samples of macromolecular complexes based on the fusion of image recognition algorithms, neural network models, and other algorithms. This algorithm can automatically determine the location of the well and the position of the target particles within it. Based on the position of the target particles within the well, the optimal imaging position is determined. According to the imaging positions, grouping is automatically optimized, and the data collection method is optimized based on these groupings. The entire process requires no manual intervention, saving significant time. Furthermore, the determined imaging positions of the target particles are more precise, and a larger number of usable particles are obtained, significantly improving the quality and quantity of data collection.

[0141] The cryo-electron microscopy data collection method of this application uses preset parsing rules to parse the medical data to be processed, and then performs data fusion on the parsed medical data to obtain fused medical data, thereby realizing the fusion of multi-source information of medical data and improving the efficiency of medical diagnosis.

[0142] The following is combined with Figure 17 The cryo-electron microscopy data collection system of this application is described, wherein, Figure 17 A schematic block diagram of a cryo-electron microscopy data collection system according to an embodiment of this application is shown.

[0143] like Figure 17 As shown, the cryo-electron microscopy data collection system 1700 includes: one or more memories 1701 and one or more processors 1702. The memories 1701 store a computer program that is executed by the processors 1702. When the computer program is executed by the processors 1702, the processors 1702 perform the cryo-electron microscopy data collection method described above.

[0144] The cryo-electron microscopy data collection system 1700 can be part or all of a computer device that can implement cryo-electron microscopy data collection methods through software, hardware, or a combination of software and hardware.

[0145] like Figure 17 As shown, the cryo-electron microscopy data acquisition system 1700 includes one or more memories 1701, one or more processors 1702, a display (not shown), and a communication interface, etc., which are interconnected via a bus system and / or other forms of connection mechanisms (not shown). It should be noted that... Figure 17 The components and structure of the cryo-electron microscopy data collection system 1700 shown are exemplary and not limiting. The cryo-electron microscopy data collection system 1700 may also have other components and structures as needed.

[0146] Memory 1701 is used to store various data and executable program instructions generated during the execution of related programs, such as algorithms for storing various application programs or implementing various specific functions. It may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0147] The processor 1702 may be a central processing unit (CPU), a graphics processing unit (GPU), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other processing units with data processing and / or instruction execution capabilities, and may be other components in the cryo-electron microscopy data collection system 1700 to perform the desired functions.

[0148] In one example, the cryo-electron microscopy data collection system 1700 also includes output devices that can output various information (such as images or sounds) to external sources (such as users), and may include one or more of a display device, a speaker, etc.

[0149] The communication interface can be any known communication protocol interface, such as a wired interface or a wireless interface. The communication interface may include one or more serial ports, USB interfaces, Ethernet ports, WiFi, wired networks, DVI interfaces, device integrated interconnect modules, or other suitable ports, interfaces, or connections.

[0150] Furthermore, according to embodiments of this application, a storage medium is also provided, on which program instructions are stored. When executed by a computer or processor, these program instructions are used to perform corresponding steps of the cryo-electron microscopy data collection method of this application. The storage medium may, for example, include a memory card of a smartphone, a storage component of a tablet computer, a hard disk of a personal computer, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a portable compact disc read-only memory (CD-ROM), a USB memory, or any combination of the above storage media.

[0151] The cryo-electron microscopy data collection system and storage medium of this application embodiment have the same advantages as the aforementioned cryo-electron microscopy data collection method because they can realize the aforementioned cryo-electron microscopy data collection method.

[0152] Although exemplary embodiments have been described herein with reference to the accompanying drawings, it should be understood that the above exemplary embodiments are merely illustrative and are not intended to limit the scope of this application. Various changes and modifications can be made therein by those skilled in the art without departing from the scope and spirit of this application. All such changes and modifications are intended to be included within the scope of this application as claimed in the appended claims.

[0153] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0154] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed.

[0155] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of this application may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0156] Similarly, it should be understood that, in order to streamline this application and aid in understanding one or more of the various inventive aspects, features of this application may sometimes be grouped together in a single embodiment, figure, or description thereof in the description of exemplary embodiments of this application. However, this approach should not be construed as reflecting an intention that the claimed application requires more features than are expressly recited in each claim. Rather, as reflected in the corresponding claims, its inventive point lies in solving the corresponding technical problem with features fewer than all features of a single disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of this application.

[0157] Those skilled in the art will understand that, apart from the mutual exclusion of features, all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or elements of any method or apparatus so disclosed can be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0158] Furthermore, those skilled in the art will understand that although some embodiments described herein include certain features but not others included in other embodiments, combinations of features from different embodiments are intended to be within the scope of this application and form different embodiments. For example, in the claims, any one of the claimed embodiments can be used in any combination.

[0159] The various component embodiments of this application can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some modules according to the embodiments of this application. This application can also be implemented as an apparatus program (e.g., a computer program and computer program product) for performing part or all of the methods described herein. Such an implementation of this application can be stored on a computer-readable medium, or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0160] It should be noted that the above embodiments are illustrative of this application and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. This application can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

[0161] The above description is merely a specific embodiment or illustration of the embodiments of this application. The scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. The scope of protection of this application shall be determined by the scope of the claims.

Claims

1. A method for collecting cryo-electron microscopy data, characterized in that, The method includes: Collect a sample map on a cryo-electron microscope grid; the sample map includes at least one grid, and each grid includes at least one circular hole; Based on the matching between at least one preset mask and the sample map, the first circular hole in the sample map that meets the preset conditions is determined; Using a preset recognition model, all target particles contained in the first circular hole are identified; Based on the positional characteristics of each target particle in the circular hole, the image data acquisition location is determined; Based on the matching between at least one preset mask and the sample map, the first circular hole in the sample map that meets preset conditions is determined, including: Each circular hole in the sample map is sequentially matched with at least one preset mask; When a match is successful, each of the circular holes is the first circular hole that meets the preset conditions; The preset mask includes at least one of the following: a complete ring, a 1 / 2 ring, and a 1 / 4 ring having a preset outer circle radius and a preset inner circle radius.

2. The method according to claim 1, characterized in that, The method further includes: The sample data is labeled and trained to construct the preset recognition model; The preset recognition model is tested to determine whether it meets the preset requirements.

3. The method according to claim 1, characterized in that, Based on the positional characteristics of each target particle within the circular hole, the image data acquisition location is determined, including: Using each target particle as the center of a single acquisition area, determine the number of target particles contained in the single acquisition area corresponding to each target particle. The area containing the largest number of target particles in a single sampling session is selected as the sampling location.

4. The method according to claim 1, characterized in that, Based on the positional characteristics of each target particle within the circular hole, the image data acquisition location is determined, including: Construct the bounding rectangle of all target particles within the first circular hole; Define a single-sampling area within the circumscribed rectangle; Within the bounding rectangle, the single-collection area is traversed and moved with a preset step size; Determine the number of target particles in the single acquisition area during each movement; The area containing the largest number of target particles in a single sampling session is selected as the sampling location.

5. The method according to claim 3 or 4, characterized in that, When at least two single-collection regions contain the maximum number of target particles, the method further includes: Determine the distance between the center of each single acquisition area and the center of the first circular hole; The sampling location is determined by identifying the area closest to the center of the circular hole in a single sampling operation.

6. The method according to claim 1, characterized in that, The sample map contains at least two first circular holes, and the method further includes: The at least two first circular holes are grouped according to a preset rule; Data is collected for each of the first circular holes in each group according to the stated collection position.

7. A cryo-electron microscopy data collection system, characterized in that, The system includes: A memory and a processor, wherein the memory stores a computer program that is executed by the processor, the computer program, when executed by the processor, causes the processor to perform the cryo-electron microscopy data collection method as described in any one of claims 1 to 6.

8. A storage medium, characterized in that, The storage medium stores a computer program that, when executed by a processor, causes the processor to perform the cryo-electron microscopy data collection method as described in any one of claims 1 to 6.

Images