Method and apparatus for generating main chain structure of protein, and device and storage medium

Abstract

A method and apparatus for generating a main chain structure of a protein, and a device and a storage medium. The method comprises: on the basis of distance information between atoms in a plurality of protein domains, generating a plurality of feature maps corresponding to the plurality of protein domains; by means of combining the plurality of feature maps, generating a fused feature map; and generating a main chain structure of a protein by means of using a first model to process the fused feature map, thereby improving the quality and stability of long-chain protein design.

Description

Method, device, apparatus and storage medium for generating backbone structure of protein TECHNICAL FIELD

[0001] Example embodiments of the present disclosure generally relate to the field of computer, and in particular, to a method, device, apparatus and computer readable storage medium for generating backbone structure of protein. BACKGROUND

[0002] With the improvement of computing power and the development of bioinformatics, it has become possible to design protein structure by computational methods. In the aspect of protein backbone design, researchers can generate protein structures with high designability by using diffusion-based protein structure generation models. However, when dealing with long-chain proteins, the design quality tends to decrease significantly as the protein sequence grows.

[0003] SUMMARY

[0004] In a first aspect of the present disclosure, a method for generating a backbone structure of a protein is provided. The method comprises: generating a plurality of feature maps corresponding to a plurality of protein domains based on distance information between atoms in the plurality of protein domains; generating a fused feature map by combining the plurality of feature maps; and generating the backbone structure of the protein by processing the fused feature map using a first model.

[0005] In a second aspect of the present disclosure, a device for generating a backbone structure of a protein is provided. The device comprises: a feature generation module configured to generate a plurality of feature maps corresponding to a plurality of protein domains based on distance information between atoms in the plurality of protein domains; a feature fusion module configured to generate a fused feature map by combining the plurality of feature maps; and a model processing module configured to generate the backbone structure of the protein by processing the fused feature map using a first model.

[0006] In a third aspect of the present disclosure, an electronic device is provided. The device comprises at least one processing unit; and at least one memory coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit. The instructions, when executed by the at least one processing unit, cause the device to perform the method of the first aspect.

[0007] In a fourth aspect of the present disclosure, a computer readable storage medium is provided. The computer readable storage medium has stored thereon a computer program executable by a processor to implement the method of the first aspect.

[0008] It should be understood that the content described in this section is not intended to limit the key features or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become apparent through the following description. BRIEF DESCRIPTION OF DRAWINGS

[0009] The above-described and other features, advantages, and aspects of the present disclosure will become more apparent as various embodiments of the present disclosure are described in greater detail below. In the drawings, like reference numerals refer to like elements throughout the various drawings, wherein:

[0010] FIG. 1 illustrates a schematic diagram of an example environment in which embodiments according to the present disclosure can be implemented;

[0011] FIG. 2 illustrates a flowchart of an example process for generating a main chain structure of a protein, according to some embodiments of the present disclosure;

[0012] FIG. 3 illustrates a schematic diagram of generating a main chain structure, according to some embodiments of the present disclosure;

[0013] FIG. 4 illustrates a schematic block diagram of an example apparatus for generating a main chain structure of a protein, according to some embodiments of the present disclosure; and

[0014] FIG. 5 illustrates a block diagram of an electronic device capable of implementing various embodiments of the present disclosure. DETAILED DESCRIPTION

[0015] Embodiments of the present disclosure will be described in more detail with reference to the drawings. While certain embodiments of the present disclosure are illustrated, it is understood that the present disclosure can be embodied in various forms and should not be interpreted as being limited to the embodiments set forth herein, but, on the contrary, these embodiments are provided for more complete and thorough comprehension of the present disclosure. It is understood that the drawings and embodiments of the present disclosure are for exemplary purposes only and are not intended to limit the scope of protection of the present disclosure.

[0016] It is noted that the headings provided herein are not limitations of the disclosure. Various embodiments are described throughout this document and any type of embodiment can be included under any heading. Additionally, embodiments described in any heading can be combined with any other embodiment described in the same heading and / or a different heading in any manner.

[0017] In the description of embodiments of the present disclosure, the term "includes" and its derivatives are to be construed as open-ended, i.e., as "including, but not limited to." The term "based on" is to be construed as "based at least in part on." The term "one embodiment" or "an embodiment" are to be construed as "at least one embodiment." The term "some embodiments" is to be construed as "at least some embodiments." Other explicit and implicit definitions can also be included below. The terms "first," "second," etc. can refer to different or same objects. Other explicit and implicit definitions can also be included below.

[0018] The data of users, acquisition and / or use of the data, etc. can be involved in the embodiments of the present disclosure. These aspects all comply with the corresponding laws and regulations and relevant provisions. In the embodiments of the present disclosure, the collection, acquisition, processing, processing, forwarding, use, etc. of all data are performed on the premise that the user is aware of and confirms. Accordingly, when implementing the embodiments of the present disclosure, the type of data or information that can be involved, the use range, the use scenario, etc. should be informed to the user and the authorization of the user should be obtained through appropriate means according to the relevant laws and regulations. The specific informing and / or authorization manner can vary according to the actual situation and application scenario, and the scope of the present disclosure is not limited in this aspect.

[0019] In the present specification and embodiments, if personal information processing is involved, it will be processed on the premise of having a legal basis (for example, obtaining the consent of the subject of personal information, or being necessary for the performance of a contract, etc.), and will only be processed within the prescribed or agreed range. The user refuses to process personal information other than the necessary information required for the basic function, which does not affect the user's use of the basic function.

[0020] As mentioned above, in the design of protein main chain, researchers can use diffusion-based protein structure generation models to generate protein structures with high designability. However, when dealing with long-chain proteins, the design quality tends to decrease significantly as the protein sequence grows.

[0021] Embodiments of the present disclosure provide a scheme for generating a main chain structure of a protein. The scheme includes: generating a plurality of feature maps corresponding to a plurality of protein domains based on distance information between atoms in the plurality of protein domains; generating a fusion feature map by combining the plurality of feature maps; and generating a main chain structure of the protein by processing the fusion feature map using a first model.

[0022] In this way, the embodiments of the present disclosure can improve the quality and stability of long-chain protein design and enhance the novelty of generated protein structures.

[0023] Various example implementations of the scheme are described in detail below in further conjunction with the accompanying drawings.

[0024] Example Environment

[0025] FIG. 1 shows a schematic diagram of an example environment 100 in which embodiments of the present disclosure can be implemented. As shown in FIG. 1, the example environment 100 can include an electronic device 110.

[0026] In this example environment 100, the electronic device 110 can be configured to generate a main chain structure 120 of a protein. The specific generation process of the main chain structure 120 will be described in detail below in conjunction with FIG. 2 and FIG. 3.

[0027] In some embodiments, the electronic device 110 can be any type of mobile terminal, fixed terminal, or portable terminal including a mobile handset, a tablet computer, a laptop computer, a notebook computer, a netbook computer, a tablet computer, a media computer, a multimedia tablet, a palmtop computer, a portable gaming terminal, a VR / AR device, a Personal Communication System (PCS) terminal, a personal navigation device, a Personal Digital Assistant (PDA), an audio / video player, a digital camera / camcorder, a positioning device, a television receiver, a radio broadcast receiver, an electronic book device, a game device, or any combinations of these, including accessories and peripherals of these devices, or any combinations thereof. In some embodiments, the electronic device 110 can also support any type of interface to a user (such as “wearable” circuitry, etc.).

[0028] The electronic device 110 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, a cloud server providing cloud service, cloud database, cloud computing, cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, content distribution network, and basic cloud computing services such as big data and artificial intelligence platform, etc. The electronic device 110 may, for example, include a computing system / server, such as a mainframe, an edge computing node, a computing device in a cloud environment, etc.

[0029] It should be understood that the structure and function of the various elements in the environment 100 are described for illustrative purposes only, without implying any limitation on the scope of the present disclosure.

[0030] Some example embodiments of the present disclosure will be further described below with reference to the accompanying drawings.

[0031] Example process

[0032] FIG. 2 shows a flowchart of an example process 200 of generating a main chain structure of a protein, according to some embodiments of the present disclosure. The process 200 can be implemented at the electronic device 110. The process 200 will be described below with reference to FIG. 1.

[0033] As shown in FIG. 2, at block 210, the electronic device 110 generates a plurality of feature maps corresponding to a plurality of protein domains based on distance information between atoms in the plurality of protein domains.

[0034] The process 200 will be further described below with reference to FIG. 3. FIG. 3 shows a schematic diagram 300 of generating a main chain structure, according to some embodiments of the present disclosure.

[0035] As shown in FIG. 3, the electronic device 110 can determine a plurality of protein domains, e.g., protein domain 310-1, protein domain 310-2, and protein domain 310-3 (collectively or individually as protein domain 310). Protein domain refers to a partial protein sequence with independent structure and function.

[0036] In some embodiments, the protein domain 310 can be a known protein domain. Alternatively or additionally, the protein domain 310 can also be generated by a protein domain generation model 305 (also referred to as a second model) as shown in FIG. 3. It should be understood that any appropriate method or model can be employed to generate the protein domain 310, examples of which can include but are not limited to FrameDiff, Chroma, and RFdiffusion, etc.

[0037] As an example, given the length of the domain, the protein domain generation model 305 (which can be represented as ) can generate a plurality of independent protein domains

[0038] Further, the electronic device 110 can generate a plurality of feature maps corresponding to the plurality of protein domains 310, e.g., feature map 315-1, feature map 315-2, and feature map 315-3 (collectively or individually as feature map 315).

[0039] In some embodiments, the feature map 315 can indicate distance information between atoms in the protein domain 310. In some scenarios, the feature map 315 can also be referred to as a distance map. As an example, corresponding to the feature map can be represented as Thus, the feature map 315 can be represented as a distance matrix, where the distance matrix can include distances between a plurality of pairs of atoms in the protein domain 310.

[0040] In block 220, the electronic device 110 generates a fused feature map 320 by combining the plurality of feature maps 315.

[0041] In some embodiments, the electronic device 110 can determine a combination order of the plurality of protein domains 310, and can process the plurality of feature maps 315 based on the combination order to generate the fused feature map 320.

[0042] As an example, the determination process of the fused feature map 320 can be represented as:

[0043] where sdm() represents a stitching operation.

[0044] As shown in equation (1), in response to a first pair of atoms in the fusion feature map 320 being associated with a same target protein domain, the electronic device 110 can determine a first distance in the fusion feature map associated with the first pair of atoms based on distance information indicated by the target protein domain. In contrast, in response to a second pair of atoms in the fusion feature map 320 being associated with different protein domains, a second distance in the fusion feature map associated with the second pair of atoms is set to a preset value (e.g., -1).

[0045] Thus, as shown in equation (1), the electronic device 110 can stitch the plurality of feature maps 310 to the diagonal part of the fusion feature map 320, and can set the non-diagonal part (representing inter-domain interaction) to -1.

[0046] At block 230, the electronic device 110 generates a main chain structure of the protein by processing the fusion feature map using a first model.

[0047] As shown in FIG. 3, the fusion feature map 320 can be further provided to a main chain structure generation model 325 (also referred to as a first model) to generate a main chain structure 120 of the protein.

[0048] Specifically, the fusion feature map 320 can be provided as edge features to guide the generation of the final protein main chain structure. The task of the main chain structure generation model 325 is to learn the pattern of deformation produced by each individual protein domain in the process of combination and the final interaction.

[0049] In some embodiments, the main chain structure generation model 325 can include a diffusion model, such as a FrameDiff model. Exemplarily, the generation process of the main chain structure 120 can be represented as:

[0050] In some embodiments, the main chain structure generation model 325 may, for example, generate a target feature map 335 based on the fusion feature map 320, and convert the target feature map 335 into the main chain structure 120. As an example, the target feature map 335 can characterize distances between pairs of atoms in the main chain structure 120.

[0051] The specific training process of the main chain structure generation model 325 will be further introduced below.

[0052] In some embodiments, the training device can segment the training main chain structure into a plurality of training protein domains, and can determine a first feature map based on the plurality of training protein domains. As an example, the training device can generate distance maps corresponding to the plurality of training protein domains based on distances between pairs of atoms in the plurality of training protein domains, and can construct the first feature map by combining the plurality of distance maps. The construction process of the first feature map may, for example, refer to the content described above with respect to equation (1).

[0053] Further, the training device can process the first feature map by using the first model to generate a predicted backbone structure. Further, the training device can determine a training loss based on a first difference between the training backbone structure and the predicted backbone structure to adjust parameters of the first model. In some embodiments, the training loss is further based on a second difference between a second feature map corresponding to the predicted backbone structure and a third feature map of the training backbone structure.

[0054] As an example, the training loss can be represented as:

[0055] In the formula (3), may indicate a difference between the first translation information of the training backbone structure and the second translation information of the predicted backbone structure; may indicate a difference between the first rotation information of the training backbone structure and the second rotation information of the predicted backbone structure; may indicate a difference between the first coordinate information of the training backbone structure and the second coordinate information of the predicted backbone structure, which can be determined by, for example, formula (4), S represents the predicted backbone structure, represents the training backbone structure; may indicate a difference between the second feature map corresponding to the predicted backbone structure and the third feature map of the training backbone structure, which can be determined by, for example, formula (5), M represents the second feature map of the training backbone structure, represents the third feature map of the training backbone structure.

[0056] In some embodiments, the backbone structure generation model 325 can also be fine-tuned based on self-play preference optimization (SPPO).

[0057] In the process of fine-tuning using SPPO, higher quality data can be represented as S w , and lower quality data can be represented as S l These data can be determined based on scTM scores, for example. These paired data are generated under the same condition M. Accordingly, the training loss used in SPPO can be represented as:

[0058] Here, π refis a copy of p^, which is kept frozen during the fine-tuning process. During the SPPO fine-tuning process, the model is made to generate more desirable results by adjusting the data towards the winner data and away from the loser data. The fine-tuning process is done by maximizing the loss function, and embodiments of the present disclosure are able to make the model generate structures similar to the winner data and significantly different from the loser data under the constraints set by the feature map (i.e., distance map) M.

[0059] Thus, embodiments of the present disclosure are not only able to design proteins with desired structural features, but also optimize the interactions between proteins, thereby providing a powerful tool for complex biological function simulation and engineering applications.

[0060] In this way, embodiments of the present disclosure are able to improve the quality and stability of long-chain protein design and enhance the novelty of generated protein structures.

[0061] Example apparatus and device

[0062] Embodiments of the present disclosure also provide a corresponding apparatus for implementing the above method or process. FIG. 4 shows a schematic structural block diagram of an example apparatus 400 for generating a main chain structure of a protein, according to certain embodiments of the present disclosure. The apparatus 400 can be implemented as or included in the electronic device 110. Various modules / components in the apparatus 400 can be implemented by hardware, software, firmware, or any combination thereof.

[0063] As shown in FIG. 4, the apparatus 400 includes: a feature generation module 410, configured to generate a plurality of feature maps corresponding to a plurality of protein domains based on distance information between atoms in the plurality of protein domains; a feature fusion module 420, configured to generate a fused feature map by combining the plurality of feature maps; and a model processing module 430, configured to generate a main chain structure of a protein by processing the fused feature map using a first model.

[0064] In some embodiments, the feature map indicates distances between a plurality of pairs of atoms in the protein domain.

[0065] In some embodiments, the feature fusion module 420 is further configured to: determine a combination order of the plurality of protein domains; and process the plurality of feature maps based on the combination order to generate the fused feature map.

[0066] In some embodiments, the apparatus 400 further includes a distance module configured to: in response to a first atom pair in the fused feature map being associated with a same target protein domain, determine a first distance associated with the first atom pair in the fused feature map based on distance information indicated by the target protein domain; and / or in response to a second atom pair in the fused feature map being associated with different protein domains, set a second distance associated with the second atom pair in the fused feature map to a preset value.

[0067] In some embodiments, the first model is trained based on a process of: segmenting the training backbone structure into a plurality of training protein domains; determining a first feature map based on the plurality of training protein domains; processing the first feature map by the first model to generate a predicted backbone structure; and determining a training loss based on a first difference between the training backbone structure and the predicted backbone structure to adjust parameters of the first model.

[0068] In some embodiments, the first difference indicates at least one of: a difference between first translation information of the training backbone structure and second translation information of the predicted backbone structure; a difference between first rotation information of the training backbone structure and second rotation information of the predicted backbone structure; and a difference between first coordinate information of the training backbone structure and second coordinate information of the predicted backbone structure.

[0069] In some embodiments, the training loss is further based on a second difference between a second feature map corresponding to the predicted backbone structure and a third feature map of the training backbone structure.

[0070] In some embodiments, the first model is further fine-tuned based on a self-play preference optimization (SPPO).

[0071] In some embodiments, the first model includes a diffusion model.

[0072] In some embodiments, the plurality of protein domains are generated by a second model.

[0073] As shown in FIG. 5, the electronic device 500 is in the form of a general electronic device. The components of the electronic device 500 can include, but are not limited to, one or more processors or processing units 510, a memory 520, a storage device 530, one or more communication units 540, one or more input devices 550, and one or more output devices 560. The processing unit 510 can be an actual or virtual processor and is capable of performing various processing according to programs stored in the memory 520. In a multi-processor system, multiple processing units perform computer-executable instructions in parallel to improve the parallel processing capability of the electronic device 500.

[0074] The electronic device 500 typically includes a plurality of computer storage media. Such media can be any available media that is accessible by the electronic device 500 and includes both volatile and nonvolatile media, removable and non-removable media. The memory 520 can be volatile (such as register, cache, RAM), non-volatile (such as ROM, EEPROM, flash memory), or some combination of the two. The storage device 530 can be a removable or non-removable media, and can include machine-readable media, such as flash drives, magnetic disks, or any other media that can be used to store information and / or data and that can be accessed by the electronic device 500.

[0075] The electronic device 500 can further include additional removable / non-removable, volatile / non-volatile storage media. Although not shown in FIG. 5, a disk drive for reading from or writing to a removable, non-volatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive for reading from or writing to a removable, non-volatile optical disk (e.g., a CD-ROM) can be provided. In such instances, each drive can be connected to the bus (not shown) by one or more data media interfaces. The memory 520 can include a computer program product 525 having one or more program modules configured to carry out the various methods or actions of the various embodiments of the present disclosure.

[0076] The communication unit 540 enables communication with other electronic devices over communication media. Additionally, the functionality of the components of the electronic device 500 can be implemented in a single computing cluster or a plurality of computer machines capable of communication over a communication connection. Thus, the electronic device 500 can operate in a networked environment using logical connections to one or more other servers, network personal computers (PCs), or another network nodes.

[0077] The input device 550 can be one or more input devices, such as a mouse, a keyboard, a trackball, etc. The output device 560 can be one or more output devices, such as a display, a speaker, a printer, etc. The electronic device 500 can also communicate with one or more external devices (not shown) such as a storage device, a display device, etc. through the communication unit 540, as needed, one or more devices that enable a user to interact with the electronic device 500, or any devices (e.g., a network card, a modem, etc.) that enable the electronic device 500 to communicate with one or more other electronic devices. Such communication can be carried out via an input / output (I / O) interface (not shown).

[0078] According to an example implementation of the present disclosure, a computer readable storage medium is provided having computer executable instructions stored thereon, where the computer executable instructions are executed by a processor to implement the method described above. According to an example implementation of the present disclosure, a computer program product is also provided that is tangibly stored on a non-transitory computer readable medium and includes computer executable instructions, where the computer executable instructions are executed by a processor to implement the method described above.

[0079] The computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0080] The computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0081] The computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0082] The computer program product of the present disclosure can have a signal including said computer program. This signal can be electronic, electromagnetic, optical, or any other suitable type of signal. Such a signal can be provided through a communication connection, such as electrical wiring, optical fiber, wireless interface, etc. Examples of computer program products include computer program implemented on a personal computer, server, or other networked device. A non-transitory computer readable medium, such as a floppy disk, CD-ROM, DVD-ROM, Blu-ray Disc, hard disk drive, or any other suitable non-transitory computer readable medium can store the computer program product.

[0083] Various implementations of the disclosure have been described in detail above. The foregoing description is exemplary and explanatory only, and is not intended to be exhaustive or to limit various implementations of the disclosure to the precise forms disclosed. Many modifications and variations are possible in light of the above teachings without departing from the scope and spirit of the disclosure. It is intended that the scope of the disclosure be limited only by the claims and the equivalents thereof. The use of the terms "including," "containing," "comprising," "having," "in involving," "portions," "elements," "components," "steps," "phases," "processes," "operations," "steps," "stages," "procedures," "methods," "mechanisms," "devices," "systems," "apparatuses," "units," "means," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "apparatuses," "units," "devices," "systems," "

Claims

1. A method for generating a main chain structure of a protein, comprising: generating a plurality of feature maps corresponding to a plurality of protein domains based on distance information between atoms in the plurality of protein domains; generating a fused feature map by combining the plurality of feature maps; and generating the main chain structure of the protein by processing the fused feature map using a first model.

2. The method of claim 1, wherein the feature map indicates distances between a plurality of pairs of atoms in a protein domain.

3. The method of claim 1, wherein generating the fused feature map by combining the plurality of feature maps comprises: determining a combination order of the plurality of protein domains; and processing the plurality of feature maps based on the combination order to generate the fused feature map.

4. The method of claim 3, further comprising: in response to a first pair of atoms in the fused feature map being associated with a same target protein domain, determining a first distance associated with the first pair of atoms in the fused feature map based on distance information indicated by the target protein domain; and / or in response to a second pair of atoms in the fused feature map being associated with different protein domains, setting a second distance associated with the second pair of atoms in the fused feature map to a preset value.

5. The method of claim 1, wherein the first model is trained based on a process comprising: segmenting a training main chain structure into a plurality of training protein domains; determining a first feature map based on the plurality of training protein domains; processing the first feature map using the first model to generate a predicted main chain structure; and determining a training loss based on a first difference between the training main chain structure and the predicted main chain structure to adjust parameters of the first model.

6. The method of claim 5, wherein the first difference indicates at least one of: a difference between first translation information of the training main chain structure and second translation information of the predicted main chain structure; a difference between first rotation information of the training main chain structure and second rotation information of the predicted main chain structure; a difference between first coordinate information of the training main chain structure and second coordinate information of the predicted main chain structure.

7. The method of claim 5, wherein the training loss is further based on a second difference between a second feature map corresponding to the predicted main chain structure and a third feature map of the training main chain structure.

8. The method of claim 1, wherein the first model is further fine-tuned based on self-play preference optimization (SPPO).

9. The method of claim 1, wherein the first model comprises a diffusion model.

10. The method of claim 1, wherein the plurality of protein domains are generated using a second model.

11. An apparatus for generating a main chain structure of a protein, comprising: a feature generation module configured to generate a plurality of feature maps corresponding to a plurality of protein domains based on distance information between atoms in the plurality of protein domains; a feature fusion module configured to generate a fused feature map by combining the plurality of feature maps; and a main chain structure generation module configured to generate the main chain structure of the protein by processing the fused feature map using a first model. ​ ​ ​ The model processing module is configured to generate a main chain structure of the protein by processing the fused feature map using a first model.

12. An electronic device, comprising: at least one processing unit; and at least one memory coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit, the instructions when executed by the at least one processing unit cause the electronic device to perform the method according to any one of claims 1-10.

13. A computer-readable storage medium having stored thereon a computer program, the computer program being executable by a processor to implement the method according to any one of claims 1-10.

Images