A protein sequence screening method, an electronic device, and a storage medium

Abstract

This application provides a protein sequence screening method, electronic device, and storage medium. The method includes: in response to receiving a sequence screening task for a target protein, acquiring a screening evidence package and candidate sequences for the target protein; the screening evidence package includes evidence data characterizing the features of the target protein, and a first mapping relationship between the residue positions of the target protein and the source of the evidence data, and the residue positions of the candidate sequences; generating and outputting a first screening rule based on the screening evidence package; in response to receiving a screening start command, screening the candidate sequences based on a second screening rule, determining and outputting the sequence screening result of the target protein; the second screening rule is determined based on the first screening rule. This application improves the screening accuracy of protein candidate sequences by automatically generating screening rules based on evidence data characterizing the features of the target protein and the residue position mapping relationship, and by screening candidate sequences based on the screening rules.

Description

Technical Field

[0001] This application relates to the field of protein engineering technology, specifically to a method for screening protein sequences, an electronic device, and a storage medium. Background Technology

[0002] With the rapid development of protein engineering technology, a vast number of candidate protein sequences can be generated through gene mutation, sequence modification, and de novo design, which greatly expands the screening range of functional proteins and provides a rich library of materials for obtaining high-performance functional proteins.

[0003] Currently, the screening of protein candidate sequences mainly relies on human experience to convert protein characteristics into screening rules. The entire process is highly dependent on personal experience, and the screening rules may be inconsistent due to differences in the cognition of different people, which cannot guarantee the accuracy of protein candidate sequence screening. Summary of the Invention

[0004] The main objective of this application is to provide a protein sequence screening method, electronic device, and storage medium, which aims to improve the screening accuracy of protein candidate sequences.

[0005] This application provides a protein sequence screening method, comprising: in response to receiving a sequence screening task for a target protein, acquiring a screening evidence package and candidate sequences for the target protein; the screening evidence package includes evidence data characterizing the features of the target protein, and a first mapping relationship between the residue positions of the target protein and the source of the evidence data, and the residue positions of the candidate sequences; generating and outputting a first screening rule based on the screening evidence package; the first screening rule is used to indicate the screening constraints that the candidate sequences need to satisfy; in response to receiving a screening start command, screening the candidate sequences based on a second screening rule, determining and outputting the sequence screening result of the target protein; the second screening rule is determined based on the first screening rule.

[0006] In one embodiment, the screening evidence package includes multiple sets of evidence data; each set of evidence data comes from a different source; the first mapping relationship includes a second mapping relationship between the target protein and the residue positions corresponding to the multiple sets of evidence data; generating and outputting a first screening rule based on the screening evidence package includes: obtaining multiple sets of residue positions of the target protein; each set of residue positions includes multiple residue positions of the target protein that have the second mapping relationship with the residue positions corresponding to the multiple sets of evidence data; for each set of residue positions, determining target evidence data based on the sub-evidence data at the corresponding residue positions of the multiple sets of evidence data; generating and outputting the first screening rule based on the target evidence data for each set of residue positions and the first mapping relationship.

[0007] In one embodiment, determining the target evidence data based on the sub-evidence data at corresponding residue positions of the multiple sets of evidence data includes at least one of the following: fusing the sub-evidence data based on the ranking of the sub-evidence data at corresponding residue positions of the multiple sets of evidence data to obtain the target evidence data; and selecting sub-evidence data matching the preset screening strategy from the sub-evidence data at corresponding residue positions of the multiple sets of evidence data in response to detecting a selection operation of a preset screening strategy to obtain the target evidence data; wherein the preset screening strategy includes at least one of a conservative screening strategy, an exploratory screening strategy, an engineering stability-priority screening strategy, and an activity retention-priority screening strategy.

[0008] In one embodiment, the evidence data includes target evidence data for at least some residue positions of the target protein; generating and outputting a first screening rule based on the screening evidence package includes: determining screening constraints corresponding to at least some residue positions of the candidate sequence based on the first mapping relationship and the target evidence data; generating and outputting the first screening rule based on the screening constraints and actions corresponding to the at least some residue positions; the actions include at least one of elimination, retention, soft penalty, weighted sorting, multi-level grouping, labeling, and risk warning.

[0009] In one embodiment, before filtering the candidate sequences based on the second filtering rule, the method further includes: In response to receiving a third filtering rule input by the user, the first filtering rule and the third filtering rule are compiled to obtain the second filtering rule; wherein the first filtering rule and the third filtering rule have the same structured fields; the compilation process includes at least one of the following: field integrity verification, residue position legality verification and mapping verification, filtering constraint expansion, strength interpretation, duplicate item merging and conflict detection, and execution order sorting.

[0010] In one embodiment, the first screening rule includes screening rules for at least one scope of action among residue sites, residue segments, site tags, structural domain segments, and the entire sequence; before screening the candidate sequence based on the second screening rule, the method further includes: in response to detecting a selection operation for a target scope of action, determining and outputting a target screening rule corresponding to the target scope of action based on the first screening rule; and determining the second screening rule based on the target screening rule.

[0011] In one embodiment, before filtering the candidate sequence based on the second filtering rule, the method further includes: in response to detecting a modification operation on the first filtering rule, determining the second filtering rule based on the modified first filtering rule; wherein the modification operation includes performing at least one of the following operations on at least a portion of the rule content in the first filtering rule: adding, editing, disabling, deleting, confirming, or copying.

[0012] In one embodiment, the sequence screening result includes at least one of the following: retained candidate sequences, eliminated candidate sequences, sorting results of candidate sequences, and grouping results of candidate sequences; after screening the candidate sequences based on the second screening rule, the method further includes: generating and outputting audit information related to the sequence screening result; the audit information includes at least one of the following: traceability information of the second screening rule, modification records of the first screening rule, and screening basis of the candidate sequences.

[0013] In one embodiment, obtaining the screening evidence package includes: obtaining first evidence data and second evidence data; the first evidence data is used to characterize the target protein in at least one dimension of three-dimensional structure and conformation features and intermolecular interactions; the second evidence data is used to characterize the target protein in at least one dimension of sequence, evolution, and function; standardizing the first evidence data and the second evidence data; and generating the screening evidence package based on the standardized first evidence data, the second evidence data, and the first mapping relationship.

[0014] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described protein sequence screening method.

[0015] This application embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the above-described protein sequence screening method.

[0016] The embodiments of this application provide a protein sequence screening method, electronic device, and storage medium that automatically generate screening rules based on evidence data characterizing the target protein, the residue positions corresponding to the source of the evidence data and the residue positions of the candidate sequence. This not only effectively avoids the problem of inconsistent screening rules caused by experience differences in traditional manual screening, but also takes into account the impact of residue position mapping relationship on screening accuracy, and can significantly reduce the risk of missed screening and false screening, thereby improving the screening accuracy of protein candidate sequences. Attached Figure Description

[0017] Figure 1 This is a schematic flowchart of the protein sequence screening method provided in the embodiments of this application.

[0018] Figure 2 This is a schematic diagram of the structure of the protein sequence screening system provided in the embodiments of this application.

[0019] Figure 3 This is a schematic diagram illustrating the specific process of the protein sequence screening method provided in the embodiments of this application.

[0020] Figure 4 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0022] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the digit " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0023] The protein sequence screening method provided in this application can be applied to electronic devices or the software of electronic devices. The electronic device can be a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet computer, laptop computer, desktop computer, etc.; the server can be configured as an independent physical server, or as a server cluster or distributed system composed of multiple physical servers. The software can be an application that implements the protein sequence screening method, but is not limited to the above forms.

[0024] The protein sequence screening method provided in this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] Please see Figure 1 The protein sequence screening method provided in this application embodiment may include: Step S101: In response to receiving a sequence screening task for the target protein, obtain the screening evidence package and candidate sequences of the target protein; the screening evidence package includes evidence data used to characterize the features of the target protein, and a first mapping relationship between the residue positions of the target protein and the source of the evidence data, and the residue positions of the candidate sequences; Optionally, the aforementioned evidentiary data may include at least one of the following: structural evidence, binding evidence, sequence evidence, family evidence, functional annotation evidence, experimental data, historical project data, local knowledge base information, or human verification information. Structural and binding evidence can be used to characterize the target protein's structure, ligand / substrate binding, pockets, interfaces, interaction networks, or conformational features, such as candidate binding poses, contact residues, first / second shell sites, hydrogen bonds, hydrophobic interactions, aromatic interactions, metal or cofactor coordination, channels, or gated regions. Sequence, family, and functional annotation evidence can be used to characterize the target protein's information regarding sequence, family, domains, conserved motifs, functional sites, active sites, binding sites, interface sites, engineering risks, or knowledge base tags.

[0026] Optionally, the source of the evidence data can be used to characterize the acquisition method of the evidence data. Considering that the reference coordinate system of the residue position of the target protein may be different under different acquisition methods, for example, for evidence data from molecular docking, structural alignment, or pocket identification, the reference coordinate system of the residue position can be the structural residue coordinate system of the target protein; for evidence data from sequence analysis, the reference coordinate system of the residue position can be the sequence coordinate system of the target protein; for evidence data from family analysis or domain analysis, the reference coordinate system of the residue position can be the family / domain coordinate system to which the target protein belongs; in addition, the residue start numbering of the candidate sequence and the target protein may be inconsistent, and the candidate sequence may have residue insertion / deletion, residue renumbering, polypeptide chain splitting, and local residue segment substitution. In practice, the first mapping relationship between the residue position of the target protein in one or more reference coordinate systems and the residue position of the candidate sequence can be established by mapping the residue position of the target protein in one or more reference coordinate systems to the reference coordinate system of the residue position of the candidate sequence. Optionally, for candidate sequences that have been inserted, missing, renumbered, split, or replaced, the first mapping relationship can be established through sequence alignment, motif anchoring, segment location, label inheritance, or manual confirmation.

[0027] Step S102: Based on the filtering evidence package, generate and output the first filtering rule; the first filtering rule is used to indicate the filtering constraints that the candidate sequence needs to satisfy. In practice, when the action is a preset action, the aforementioned screening constraints can be determined based on the evidence data, and the aforementioned first screening rule can be directly determined based on the determined screening constraints and the aforementioned first mapping relationship. When the action is not a preset action, the aforementioned screening constraints can be determined first based on the evidence data, and the action can be determined based on the level of the evidence data. For specific implementation details, please refer to the relevant descriptions below, which will not be described here. Then, the aforementioned first screening rule is generated based on the aforementioned screening constraints, the aforementioned first mapping relationship, and the action. Optionally, when a candidate sequence meets the aforementioned screening constraints, the preset action can be to retain the candidate sequence; when a candidate sequence does not meet the aforementioned screening constraints, the preset action can be to eliminate the candidate sequence.

[0028] Since the first mapping relationship can be used to characterize the correspondence between the residue positions of the evidence data in its source coordinate system, the residue positions in the target protein reference coordinate system, and the residue positions in the candidate sequence coordinate system, the first screening rule generated based on the evidence data and the first mapping relationship can be used to indicate the screening constraints that the candidate sequence should satisfy at the first residue position or within the region to which the first residue position belongs. The first residue position can be any residue position in the candidate sequence.

[0029] The screening constraints can be determined based on structural evidence, binding evidence, sequence evidence, family evidence, functional annotation evidence, experimental data, and / or human verification information of the target region corresponding to the first residue position in the target protein. The target region may include the second residue position corresponding to the first residue position in the target protein reference coordinate system, or it may include at least one of the following: a residue segment, a domain segment, a functional site region, a binding pocket region, a site tag region, or a whole sequence region containing the second residue position.

[0030] Step S103: In response to receiving the screening start command, the candidate sequences are screened based on the second screening rule, and the sequence screening results of the target protein are determined and output; the second screening rule is determined based on the first screening rule.

[0031] Optionally, the first filtering rule can be directly determined as the second filtering rule, or the second filtering rule can be determined based on the first filtering rule, combined with the third filtering rule input by the user and / or operation information related to the first filtering rule. For specific implementation details, please refer to the relevant descriptions below, which will not be described here.

[0032] In practice, for each residue position in at least a subset of the candidate sequence, a second screening rule can be used to determine the screening constraints that the candidate sequence should satisfy at that residue position or within the region containing that residue position. Then, the characteristics of the candidate sequence at that residue position or within the region containing that residue position can be compared with the corresponding screening constraints to determine whether the candidate sequence satisfies the screening constraints at that residue position or within the region containing that residue position. Finally, based on whether the candidate sequence satisfies the screening constraints at each residue position or within each residue position's region, corresponding processing actions can be performed on the candidate sequence to obtain the sequence screening results of the target protein.

[0033] This application embodiment automatically generates screening rules based on evidence data characterizing the target protein, the first mapping relationship between the residue positions of the target protein and the source of the evidence data, and the residue positions of the candidate sequence. This not only effectively avoids the problem of inconsistent screening rules caused by experience differences in traditional manual screening, but also takes into account the impact of residue position mapping relationship on screening accuracy, and can significantly reduce the risk of missed screening and false screening, thereby improving the screening accuracy of protein candidate sequences.

[0034] Optionally, the aforementioned screening evidence package may include multiple sets of evidence data, each set of evidence data having a different source; in this case, each set of evidence data may include at least one sub-evidence data; each sub-evidence data may be used to characterize the features of the target protein at the residue position corresponding to the sub-evidence data; the aforementioned first mapping relationship may include a second mapping relationship between the target protein and the residue positions corresponding to the multiple sets of evidence data respectively.

[0035] In one embodiment, the step S102 above, which generates and outputs a first screening rule based on the screening evidence package, includes: Obtain multiple sets of residue positions of the target protein; each set of residue positions includes multiple residue positions of the target protein that have a second mapping relationship with the residue positions corresponding to multiple sets of evidence data; For each group of residue positions, target evidence data is determined based on sub-evidence data at the corresponding residue positions from multiple groups of evidence data. Based on the target evidence data of each group of residue positions and the first mapping relationship, the first screening rule is generated and output.

[0036] Optionally, the second mapping relationship described above can be used to group multiple residue positions that have a second mapping relationship with the residue positions corresponding to multiple sets of evidence data into a group, thus obtaining multiple sets of residue positions of the target protein. In actual implementation, multiple residue positions from different evidence data sources that correspond to the same reference residue site or the same reference residue segment can be grouped into a group according to the second mapping relationship, thus obtaining multiple sets of aligned residue positions, i.e., the multiple sets of residue positions of the target protein. In this way, each set of residue positions may include one residue position of the target protein corresponding to each of the multiple sets of evidence data, and the second mapping relationship exists between the residue positions in the same group. Optionally, the above residue positions can be represented by at least one of the following: a single residue site, a residue segment composed of multiple consecutive residue sites, a domain segment, or the entire residue sequence (also known as the whole sequence). Then, the target evidence data can be determined based on the sub-evidence data at the corresponding residue positions of the multiple sets of evidence data. For specific implementation, please refer to the relevant description below, which will not be described here. Next, based on the target evidence data for each group of residue positions and the aforementioned first mapping relationship, sub-screening rules corresponding to that group of residue positions can be determined; finally, the aforementioned first screening rule can be determined based on the sub-screening rules corresponding to each group of residue positions. Optionally, the first screening rule may include the sub-screening rules corresponding to each group of residue positions, or it may include the screening rule obtained by integrating the sub-screening rules corresponding to each group of residue positions.

[0037] This application embodiment obtains multiple sets of residue positions of the target protein; and for each set of residue positions, determines the target evidence data based on the sub-evidence data at the corresponding residue positions of multiple sets of evidence data; and generates and outputs a first screening rule based on the target evidence data of each set of residue positions and a first mapping relationship. This can ensure the reliability of the first screening rule generated under multiple sets of evidence data, thereby improving the accuracy of screening candidate sequences using the first screening rule.

[0038] In one embodiment, the determination of target evidence data based on sub-evidence data at corresponding residue positions from multiple sets of evidence data includes at least one of the following: Based on the ranking of sub-evidence data at corresponding residue positions of multiple sets of evidence data, the sub-evidence data are fused to obtain the target evidence data. In response to the detection of a selection operation of a preset screening strategy, sub-evidence data matching the preset screening strategy is selected from sub-evidence data at the corresponding residue positions of multiple sets of evidence data to obtain target evidence data; the preset screening strategy includes at least one of a conservative screening strategy, an exploratory screening strategy, an engineering stability-priority screening strategy, and an activity retention-priority screening strategy.

[0039] Optionally, based on the confidence and / or importance of sub-evidence data at corresponding residue positions from multiple sets of evidence data, the evidence data can be stratified to obtain the level of each sub-evidence data. High-level evidence data may include clearly defined functional sites, clearly defined catalytic / binding / coordination sites, strongly conserved motif core residues, or important sites confirmed by human intervention; medium-level evidence data may include sites where structural contact evidence overlaps with family evidence, key gating loop sites, interface hotspots, or important segments; low-level evidence data may include candidate sites obtained solely from single structural analysis, single model hints, or empirical statistics.

[0040] In practice, when dealing with sub-evidence data at corresponding residue positions from multiple sets of evidence data, conflicts can be identified when the screening constraints, allowed / prohibited residue sets, physicochemical property directions, treatment actions, or priorities indicated by the multiple sub-evidence data are inconsistent. In this case, the sub-evidence data with the highest priority can be selected as the target evidence data. Alternatively, when multiple sub-evidence data are consistent or complementary, they can be fused to obtain the target evidence data. For example, the single piece of evidence data with the most complete information dimension can be selected from multiple consistent evidence data as the target evidence data. Or, multiple complementary pieces of evidence can be jointly identified as the target evidence data.

[0041] In practice, if the user selects a preset screening strategy, the target evidence data can be obtained by selecting sub-evidence data that matches the preset screening strategy from sub-evidence data at corresponding residue positions in multiple sets of evidence data, based on the preset matching relationship between the screening strategy and the evidence data level. Optionally, after selecting the sub-evidence data that matches the preset screening strategy, the selected sub-evidence data can be further fused to obtain the target evidence data. For specific implementation details, please refer to the relevant descriptions above, which will not be repeated here.

[0042] Optionally, a conservative screening strategy can be used to indicate the use of screening rules based on fixed identity, motif matching, or strong confinement corresponding to high-level evidence data; an exploratory screening strategy can be used to indicate the use of more lenient screening rules derived from the transformation of some medium- and low-level evidence data, such as physicochemical constraints, soft penalties, or suggestiveness; an engineering stability-first screening strategy can be used to indicate the use of screening rules that place greater emphasis on whole-sequence statistics, disulfide bond risk, extreme charge changes, or segmental mutation budgets; and an activity retention-first screening strategy can be used to indicate the use of screening rules that place greater emphasis on activity, binding, catalysis, or pocket-related evidence.

[0043] This application embodiment obtains target evidence data by fusing sub-evidence data at corresponding residue positions of multiple sets of evidence data based on rank, and / or by selecting sub-evidence data that matches a preset screening strategy from the sub-evidence data at corresponding residue positions of multiple sets of evidence data. This can ensure the flexibility and reliability of determining target evidence data under multiple sets of evidence data, and enable the use of target evidence data to improve the screening accuracy of candidate sequences.

[0044] Optionally, the aforementioned evidence data includes target evidence data covering at least some residue positions of the target protein. In one embodiment, the step S102 above, which generates and outputs a first screening rule based on the screening evidence package, includes: For at least some residue positions of a candidate sequence, based on a first mapping relationship and target evidence data, the screening constraints corresponding to the residue positions are determined; Based on the screening constraints and actions corresponding to at least some residue positions, a first screening rule is generated and output; the actions include at least one of elimination, retention, soft penalty, weighted sorting, multi-level grouping, labeling, and risk warning.

[0045] In practice, when the screening evidence package includes a set of evidence data, the set of evidence data may include sub-evidence data of at least some residue positions of the target protein. In this case, for each residue position of at least some residue positions of the target protein, the sub-evidence data of that residue position can be determined as the target evidence data of that residue position.

[0046] When the evidence package includes multiple sets of evidence data, each residue position in at least some residue positions of the target protein can be used to characterize any residue position in a set of residue positions corresponding to the multiple sets of evidence data. The target evidence data for that residue position can be the target evidence data for the set of residue positions in which that residue position is located. The process for determining each set of residue positions corresponding to the multiple sets of evidence data and the target evidence data for each set of residue positions can be found in the description in the above embodiments, and will not be repeated here.

[0047] In practical implementation, for each residue position in at least a portion of the residue positions of the candidate sequence, target evidence data corresponding to that residue position can first be determined based on the aforementioned first mapping relationship and the target evidence data for at least a portion of the residue positions of the target protein. Then, based on the characteristics of the target protein represented by the target evidence data corresponding to that residue position, screening constraints corresponding to that residue position can be extracted and generated. Next, the corresponding action for that residue position can be determined based on at least one of the type, level, confidence level, preset screening strategy, and user configuration of the target evidence data. Optionally, the action matching high-level evidence data can be elimination or retention; the action matching medium-level evidence data can include at least one of soft penalty, weighted ranking, and multi-level grouping; the action matching low-level evidence data can include at least one of labeling and risk warning. Optionally, multi-level grouping can be used to indicate the grouping of candidate sequences based on the matching of the characteristics of the candidate sequence at each residue position with the characteristics of the target protein at the corresponding residue positions.

[0048] Then, for each residue position in at least a subset of the residue positions in the candidate sequence, a sub-screening rule can be generated based on the screening constraints and actions corresponding to that residue position. Next, the sub-screening rule for each residue position can be determined as the first screening rule described above. Alternatively, the sub-screening rules for each residue position can be integrated to obtain at least one screening rule, and then the integrated at least one screening rule can be determined as the first screening rule described above. It can be understood that the first screening rule can be represented as a screening rule package, which may include the sub-screening rules for each residue position or the integrated at least one screening rule.

[0049] In practice, the sub-screening rules corresponding to the same disposal action can be integrated into a single screening rule based on the disposal action indicated by the sub-screening rules at at least some residue positions of the candidate sequence; or the sub-screening rules with the same scope of action can be integrated into a single screening rule based on the scope of action corresponding to each residue position. This makes it easier to arrange the execution order of each screening rule and improve the screening efficiency of the candidate sequence.

[0050] Optionally, if the target evidence data corresponding to a certain residue position in a candidate sequence includes multiple sub-evidence data of the same level, a more lenient physicochemical category rule, a soft penalty rule, or a suggestive sub-screening rule can be generated for that residue position. Optionally, by further combining with a preset screening strategy, screening rules as shown in Table 1 can be generated: Table 1 Examples of Filtering Rules

[0051] Optionally, for different preset screening strategies, the generated first screening rule can be a strategy rule package corresponding to the preset screening strategy, such as a conservative strategy rule package, a balanced strategy rule package, an exploratory strategy rule package, an engineering stability priority strategy rule package, an activity preservation priority strategy rule package, or a user-defined strategy rule package. The conservative strategy rule package may include screening rules based on fixed identity, motif matching, or strong confinement corresponding to high-confidence evidence; the exploratory strategy rule package may include more lenient physicochemical constraints, soft penalties, or suggestive screening rules derived from some medium- and low-confidence evidence; the engineering stability priority strategy rule package may include screening rules that place greater emphasis on whole-sequence statistics, disulfide bond risk, extreme charge changes, or segmental mutation budgets; and the activity preservation priority strategy rule package may include screening rules that place greater emphasis on activity, binding, catalysis, or pocket-related evidence. Thus, the screening method provided by this application embodiment allows for flexible switching between strategy rule packages based on activity preservation, engineering stability, exploration space, and user project preferences, thereby improving the accuracy of screening candidate sequences using strategy rule packages.

[0052] This application embodiment determines screening constraints corresponding to at least some residue positions of a candidate sequence based on a first mapping relationship and target evidence data; determines disposal actions corresponding to residue positions based on the level of the first mapping relationship and target evidence data; and generates and outputs a first screening rule based on the screening constraints and disposal actions corresponding to at least some residue positions. This ensures the reliability of the generated first screening rule and improves the accuracy of screening candidate sequences using the first screening rule.

[0053] In one embodiment, before filtering the candidate sequences based on the second filtering rule in step S103 above, the filtering method provided in this application embodiment further includes: In response to receiving the third filtering rule input by the user, the first filtering rule and the third filtering rule are compiled and processed to obtain the second filtering rule; The first and third filtering rules have the same structured fields; the compilation process includes at least one of the following: field integrity verification, residue position validity verification and mapping verification, filtering constraint expansion, strength interpretation, duplicate item merging and conflict detection, and execution order sorting.

[0054] In practice, when a first filtering rule is automatically generated based on the filtering evidence package, the generated first filtering rule can be displayed in the filtering rule configuration interface. Next, the user can input a third filtering rule for filtering candidate sequences through the filtering rule configuration interface; then, in response to receiving the user-input third filtering rule, the above-mentioned compilation process can be performed on the first and third filtering rules to obtain the second filtering rule, and the finalized second filtering rule can be displayed on the filtering rule configuration interface.

[0055] Optionally, the expansion of the aforementioned screening constraints may include at least one of physicochemical template expansion and sequence and segment rule expansion; the expression of the intensity interpretation of the aforementioned screening constraints may be a directional tiered interpretation. The aforementioned execution order sorting may include at least one of the following: sorting the execution order of the first screening rule and the second screening rule; sorting the multiple screening rules included in the first screening rule and / or the second screening rule.

[0056] In practice, strong constraint rules for elimination or retention can be executed first, followed by flexible handling rules such as soft penalties, weighted sorting, and multi-level grouping. This approach aims to improve the efficiency of candidate sequence selection while ensuring the accuracy of the selection process. Optionally, the structured fields of the selection rules can be as shown in Table 2. Table 2. Structured fields of the filtering rules

[0057] Optionally, when the action involves soft penalties, weighted sorting, or score adjustments, the field value may also include a penalty parameter, a weight parameter, or a sorting parameter.

[0058] In this embodiment, in response to receiving a third filtering rule input by the user, the first filtering rule and the third filtering rule are compiled to obtain a second filtering rule. The first filtering rule and the third filtering rule have the same structured fields, which can realize the efficient fusion and reuse of automatic rules and manual rules, and can further improve the reliability of the determined filtering rules, so that the filtering accuracy of candidate sequences can be improved by using the second filtering rule.

[0059] Optionally, the first screening rule mentioned above includes screening rules targeting at least one scope of action among residue sites, residue segments, site tags, structural domain segments, and the entire sequence. In one embodiment, before screening candidate sequences based on the second screening rule in step S103 above, the screening method provided in this application embodiment further includes: In response to the detection of a selection operation targeting the scope of the target, the target filtering rule corresponding to the scope of the target is determined and output based on the first filtering rule. Based on the target screening rules, the second screening rules are determined.

[0060] In practice, a filtering rule configuration interface can be displayed to the user, showing the structured fields and their values ​​for the first filtering rule. Optionally, for the scope field, its value can include at least one of residue site, residue segment, site tag, domain segment, and whole sequence; the user can select the target scope by choosing the specific field value of the scope field.

[0061] Optionally, the first filtering rule may include at least one filtering rule with a scope. Upon detecting a selection operation targeting a specific scope, the target filtering rule corresponding to the target scope can be extracted from the first filtering rule and displayed in the filtering rule configuration interface.

[0062] The target filtering rule can then be directly determined as the second filtering rule; alternatively, the second filtering rule can be determined by further combining the third filtering rule input by the user and / or the modification information related to the target filtering rule. For specific implementation details, please refer to the description in the relevant embodiments, which will not be repeated here.

[0063] Optionally, the above target filtering rules may include at least one of the following: Site identification rules: used to identify functional sites, key catalytic / binding / coordination sites, and important sites confirmed by humans.

[0064] Site allow / prohibit residue rules: used for a specific site, a group of sites, a range, a domain, or a tag segment.

[0065] Site physicochemical property rules: for example, it must be hydrophobic, must not be positively charged, must be an aromatic residue, and must not introduce specific risky residues, etc.

[0066] Relative site variation rules: For example, relative to the reference sequence, the side chain volume must not increase, hydrophobicity must not decrease, charge must not be flipped, and aromaticity must be maintained.

[0067] Motif or segment rules: For example, key motifs must remain matched, core sites must not be destroyed, and peripheral sites can vary within a limited range.

[0068] Counting and budgeting rules: such as the maximum number of mutations in the entire sequence, the maximum number of mutations in a certain segment, the upper limit of the number of new residues of a certain type, and the upper limit of the number of changes in a certain property.

[0069] Full sequence statistical rules: such as control of the number of newly added cysteine ​​residues or potential disulfide bond risks, upper limit of changes in charged residues, range of hydrophobic residue ratios, and risk warnings for low-complexity fragments or abnormal compositions.

[0070] Thus, the screening method provided by this scheme can handle both limited site modification candidate sequence screening and large-scale sequence redesign candidate sequence screening. Furthermore, during the screening process, it can check whether the mutation is reasonable and whether the entire candidate sequence still meets functional, structural, and engineering constraints.

[0071] In this embodiment, in response to the detection of a selection operation targeting a target range, a target filtering rule corresponding to the target range is determined and output based on a first filtering rule; and a second filtering rule is determined based on the target filtering rule. This not only improves the targeting of the determined second filtering rule, but also adapts to diverse filtering needs and enhances the flexibility of filtering rule configuration, thereby improving the accuracy of filtering candidate sequences using the second filtering rule.

[0072] In one embodiment, before filtering the candidate sequences based on the second filtering rule in step S103 above, the filtering method provided in this application embodiment further includes: In response to the detection of a modification operation against the first filtering rule, a second filtering rule is determined based on the modified first filtering rule; The modification operation includes performing at least one of the following operations on at least a portion of the rule content in the first filtering rule: adding, editing, disabling, deleting, confirming, or copying.

[0073] In practice, the filter rule configuration interface can provide users with controls that allow modification of the first filter rule. Users can modify at least a portion of the rule content in the first filter rule by operating the corresponding controls.

[0074] Optionally, the rule content of the first filtering rule may include the filtering rules included in the first filtering rule, and the field values ​​of the structured fields of individual filtering rules. In actual implementation, filtering rules can be added to the first filtering rule, and / or at least some filtering rules in the first filtering rule can be disabled / deleted, and / or the field values ​​of the structured fields of at least some filtering rules in the first filtering rule can be edited.

[0075] The modified first filtering rule can then be determined as the second filtering rule. Alternatively, the modified first filtering rule can be further combined with the third filtering rule input by the user and / or the target filtering rule determined by the selection operation based on the target scope to determine the second filtering rule. For specific implementation details, please refer to the description in the relevant embodiments, which will not be repeated here.

[0076] This application embodiment, in response to the detection of a modification operation on the first screening rule, determines a second screening rule based on the modified first screening rule, which can realize the dynamic adjustment of the screening rule and ensure the accuracy of screening candidate sequences using the finally determined second screening rule.

[0077] Optionally, the above sequence screening results may include at least one of the following: retained candidate sequences, eliminated candidate sequences, sorting results of candidate sequences, and grouping results of candidate sequences.

[0078] Where the aforementioned actions include labeling and / or risk warnings, the sequence screening results may also include labeling and / or risk warnings related to the candidate sequences.

[0079] In one embodiment, after filtering the candidate sequences based on the second filtering rule in step S103 above, the filtering method provided in this application embodiment further includes: Generate and output audit information related to the sequence screening results; the audit information includes at least one of the following: traceability information of the second screening rule, modification records of the first screening rule, and screening basis of the candidate sequences.

[0080] In practice, the source information of the second screening rule mentioned above may include the rule source, rule compilation version, evidence package version, automatic rule generation result, operator information and execution time; the modification record of the first screening rule may include the content manually modified by the user, the information of the modifier and the modification time; the screening criteria for candidate sequences may include the rule number and rule description violated by each candidate sequence, and details of soft penalty / weighted sorting / multi-level grouping.

[0081] This application embodiment retains three types of data in a hierarchical manner: traceability information of the second screening rule, modification records of the first screening rule, and candidate sequence screening basis. Relying on the relationship between the second screening rule and the first screening rule, the entire process from basic rule revision and rule compilation to sequence screening determination can be completely traced back, realizing a fully closed-loop and auditable screening process. At the same time, the retained data can be packaged into templates for reuse, which is convenient for project review, training, and subsequent R&D management.

[0082] In one embodiment, obtaining the screening evidence package in step S101 includes: Acquire first evidence data and second evidence data; the first evidence data is used to characterize the target protein in at least one dimension of three-dimensional structure and conformation features and intermolecular interactions; the second evidence data is used to characterize the target protein in at least one dimension of sequence, evolution, and function. Standardize the primary and secondary evidence data; Based on the standardized first and second evidence data and the first mapping relationship, a screening evidence package is generated.

[0083] Optionally, the primary evidence data can originate from the results of molecular docking, structure alignment, pocket identification, interaction analysis, conformational ensemble analysis, experimental structures, simulations, or manual verification. In practice, results from different sources can be uniformly abstracted, retaining at least some of the following fields: evidence number, source type, target protein chain identifier, residue number, spatial region, score or confidence level, associated ligand / substrate or pocket, interaction type, and evidence description.

[0084] The values ​​of at least some of the above fields can be determined by performing the following operations: extracting contact residues related to ligands, substrates, cofactors, metals, interfaces, or pockets according to a preset distance threshold or interaction criterion; identifying hydrogen bonds, salt bridges, hydrophobic interactions, aromatic interactions, metal or cofactor coordination, steric repulsion, channel bottlenecks, or gated features; and identifying first shell sites, second shell sites, interface hotspots, channel / gated regions, or pocket support regions based on distance, interaction strength, and spatial hierarchy. Next, at least one of the following processing methods—union, intersection, or weighted union—can be performed on the contact residue set, site set, segment set, or evidence record set to obtain standardized first evidence data. Then, consensus frequency statistics or confidence normalization processing can be performed on the first evidence data. At the same time, structural residue numbers can be mapped to reference sequence positions, and mapping confidence and potential anomalies can be recorded to establish the aforementioned first mapping relationship.

[0085] Optionally, the second evidence data can be derived from family analysis, domain analysis, conservation analysis, functional site annotation, database annotation, local knowledge base, historical experimental database, literature review, or manually verified data. In practice, at least some data can be extracted from the target protein's family entries, domain segments, repeat segments, important sites, conserved motifs, catalytically related sites, binding-related sites, metal or cofactor-related sites, interface hotspots, gating loops or channel-related sites, engineering risk tags, and other functional tags to obtain the aforementioned second evidence data.

[0086] Next, the aforementioned second piece of evidence data can be standardized: site, segment, motif, and textual evidence can be converted into structured fields; for each piece of evidence, the source type, source identifier, evidence level, priority, and confidence level can be recorded. If necessary, family coordinates, domain coordinates, or motif coordinates can be mapped to reference sequences and structural residue numbers to establish the aforementioned first mapping relationship; additionally, evidence that cannot be reliably mapped can be marked to avoid directly generating strong constraint rules.

[0087] Optionally, the aforementioned screening evidence package may include at least some of the following fields: target protein identifier, reference sequence, chain identifier, list of structural residues, list of structural and binding evidence, list of sequence / family / functional annotation evidence, site tag, segment tag, coordinate mapping table, evidence source, evidence confidence level, evidence version, and generation time. In actual implementation, the field values ​​of the screening evidence package can be determined based on the standardized first evidence data and second evidence data, and the aforementioned position / coordinate mapping results.

[0088] This application embodiment obtains first evidence data for characterizing the target protein in at least one dimension of three-dimensional structure and conformation features and intermolecular interactions, and second evidence data for characterizing the target protein in at least one dimension of sequence, evolution, and function; standardizes the first evidence data and the second evidence data; and generates a screening evidence package based on the standardized first evidence data, the second evidence data, and a first mapping relationship. This improves the reliability of the generated screening evidence package, thereby improving the accuracy of screening candidate sequences using the screening evidence package.

[0089] The screening method provided in this application embodiment can be applied to, for example, Figure 2 The protein sequence screening system shown may include: an upstream evidence receiving module, an evidence standardization module, an evidence package construction and coordinate mapping module, an evidence stratification and conflict resolution module, a strategy rule package generation module, a rational screening configuration module, a rule compilation module, a rule execution module, a result output and auditing module, and a storage module.

[0090] Among them, the upstream evidence receiving module can be used to receive upstream evidence data such as structure, combination, sequence, family, functional annotation, experimental data, knowledge base, and manual confirmation information; The evidence standardization module can be used to convert evidence data from different sources into uniform fields, including evidence data type, location, segment, source, confidence level, version, and description; The evidence package construction and coordinate mapping module can be used to establish the mapping relationship between structural residue coordinates, reference sequence coordinates, candidate sequence coordinates, family / domain coordinates and motif coordinates, and combine them with upstream evidence data to form a screening evidence package; The evidence stratification and conflict resolution module can be used to stratify evidence data, determine priorities, identify conflicts, and handle conflicts. The strategy rule package generation module can be used to automatically generate one or more filtering rules based on the filtering evidence package and preset / custom filtering strategies, and can be organized into one or more strategy rule packages; The rational filtering configuration module can be used to display automatically generated filtering rules and supports users to manually add, edit, disable, delete, confirm or copy filtering rules; The rule compilation module can be used to convert automatic and manual filtering rules into executable rule objects, and perform validation, expansion, merging and sorting; optionally, the rule object can be a structured representation of an executable or compilable filtering rule, which may include structured fields and their field values; The rule execution module can be used to eliminate, retain, reorder, label, or provide risk warnings to candidate sequences based on filtering rules; The results output and audit module can be used to output filtering results, violation details, rule sources, manual modification records, execution logs, and audit records. The storage module can be used to save filtering evidence packages, rule objects, policy rule packages, execution logs, project templates, and filtering results.

[0091] Optionally, the policy rule package generation module can employ a template engine, machine learning model, text model, rule base, knowledge graph, or a combination thereof. Instead of directly performing a final black-box judgment on candidate sequences, the policy rule package generation module can automatically generate structured rule objects or candidate rules based on the filtered evidence package, and then the rule compilation module and rule execution module can perform deterministic implementation. This allows for the simultaneous utilization of automated information processing capabilities and deterministic rule execution capabilities. In practice, this can be implemented in two stages: first, candidate rules are generated by the template engine, rule base, and / or artificial intelligence model; then, a normalizer uniformly converts them into rule objects; and finally, the rule compiler performs field validation, conflict checking, and executable processing.

[0092] Please combine Figure 2 and Figure 3 In one specific embodiment, the screening method provided in this application may further include, but is not limited to, the following steps: S10: Receive upstream evidence data and candidate sequence set for the target protein; the upstream evidence data may come from structural analysis, binding analysis, pocket analysis, sequence analysis, family analysis, conservation analysis, functional annotation, experimental data or knowledge base.

[0093] S20: Standardizes upstream structure and binding evidence; can extract information such as pockets, candidate binding conformations, scores or confidence levels, contact residues, interaction types, first shell sites, second shell sites, interface regions or gating regions.

[0094] S30: Standardize upstream sequence, family and functional annotation evidence; extract family entries, structural domain segments, important sites, conserved motifs, functional labels, source information and confidence levels.

[0095] S40: Establish positional mapping relationships; it is possible to establish mapping relationships between structural residue positions, reference sequence positions, candidate sequence positions, family / domain coordinate positions, and tag segments, and combine them with standardized evidence data to form a screening evidence package; the candidate sequence set can be used as input for the screening task, but is not a necessary component of the screening evidence package.

[0096] S50: Based on the filtered evidence package, perform evidence stratification, priority judgment and conflict resolution to generate one or more filtering rules; optionally, it can combine preset or user-defined filtering strategies to obtain one or more strategy rule packages, such as conservative rule package, balanced rule package, exploratory rule package, engineering stability priority rule package, activity preservation priority rule package or user-defined strategy rule package.

[0097] S60: Automatically generated policy rules are populated back into the rational filtering configuration module to support users in viewing, confirming, enabling, disabling, editing, deleting, or adding manual rules.

[0098] S70: Integrates automatic and manual filtering rules; it can compile automatically generated and manually entered filtering rules into an executable rule object and perform field validation, template expansion, coordinate mapping validation, conflict checking, and execution order sorting.

[0099] S80: Perform filtering, penalty, reordering, labeling, or risk warning on candidate sequences based on the fused filtering rules.

[0100] S90: Outputs the sequence screening results of the target protein and related audit information; it can output retained candidate sequences, eliminated candidate sequences, violation details, rule sources, manual modification records, final sorting results and audit records, and can save them as project templates or rule templates for subsequent reuse.

[0101] The specific implementation process of this embodiment can be found in the relevant descriptions in the above embodiments, and will not be repeated here.

[0102] Optionally, the rational filtering configuration module can adopt a unified rule configuration method as the human-computer interaction platform. This configuration method can be used to display automatically generated filtering rules, as well as to accommodate filtering rules that users manually add, edit, disable, or confirm. Automatic and manual filtering rules can use the same rule object structure, allowing automatically generated filtering rules to be directly populated back into the rational filtering configuration module's filtering rule configuration interface for users to confirm and modify one by one. Users can also add filtering rules derived from engineering experience, literature, or experimental summaries, or project-defined filtering rules, within the same configuration interface.

[0103] Optionally, the rational filtering configuration module may include a filtering switch area, a policy rule package selection area, a filtering rule configuration area, a filtering rule description area, and an audit information area. Each row or item in the filtering rule configuration area can correspond to an independent filtering rule, and multiple filtering rules can be executed according to automatic sorting, user-defined order, priority, or execution action order.

[0104] To balance ease of use and program executability, the above-mentioned screening rule configuration area can support the scope of action of single site, multiple sites, segments, all mutation sites, all sites and the whole sequence; it can support trigger conditions such as checking only when mutation occurs, checking all sites, checking segment statistical features and checking the statistical features of the whole sequence; it can support screening constraint types such as allowed residues, prohibited residues, physicochemical properties, relative changes, motifs / segments and count budgets; and it can support disposal actions such as elimination, retention, soft penalty, weighted sorting, labeling or risk warning.

[0105] The protein sequence screening method provided in this application is applicable to post-processing screening scenarios of candidate sequences output by protein artificial intelligence design models and various protein sequence generation and optimization methods. It can be seamlessly embedded into existing protein design workflows or deployed independently in a modular fashion. It is adaptable to upstream evidence systems output by multi-dimensional pre-processing analysis and flexibly adapts to post-processing business workflows for sequence screening. Furthermore, the screening method provided in this application innovatively integrates a user-oriented structured rational screening module with an automatic rule generation mechanism, balancing machine-automated rule recommendation with manual custom configuration capabilities. This enables intelligent, structured, and controllable implementation of post-processing screening of protein candidate sequences. In addition, the screening method provided in this application also has the following beneficial effects: By automatically transforming a multi-dimensional evidence system, including upstream structure, binding properties, sequence characteristics, family conservation, and functional annotation, into standardized rational screening rules, the workload of manual evidence compilation and rule summarization can be significantly reduced. This effectively avoids the problem of missing key evidence due to manual analysis and significantly improves the efficiency and completeness of screening rule construction. Simultaneously, relying on positional mapping relationships, precise matching of structural sites, family structural domain sites, functional annotation sites, and candidate sequence sites can be achieved. This reduces the risk of mismatched coordinates among multiple systems and ensures the accuracy of screening candidate sequences using screening rules.

[0106] A unified rule object system has been constructed, enabling standardized management of both automatically generated and manually configured rules. Through standardized rule compilation and deterministic execution mechanisms, the sequence screening process possesses interpretable, auditable, and replayable technical characteristics. In terms of rule adaptation, it supports single-point, multi-point, functional segment, tag segment, and whole-sequence statistical screening rules, compatible with various protein modification scenarios such as precise modification of limited sites and large-scale sequence redesign. It adapts to the screening needs of protein candidate sequences from various sources, exhibiting strong versatility and scenario adaptability.

[0107] It supports the configuration of multiple differentiated strategy rule packages, allowing for flexible switching of screening strategies based on different objectives such as activity preservation, engineering stability optimization, sequence exploration space expansion, and personalized project preferences, adapting to diverse screening scenario needs. Simultaneously, it achieves full-process information traceability, including rule generation source, upstream evidence version, manual modification records, and sequence violation details, forming a complete screening process traceability system. This provides comprehensive data support for project result verification, R&D personnel training, reuse of high-quality rule templates, and subsequent iterative R&D management, effectively improving the standardization and normalization of protein sequence screening.

[0108] This application also provides an electronic device, including a processor and a memory. The memory stores a program or instructions that can be executed on the processor. When the program or instructions are executed by the processor, they implement the various steps of the protein sequence screening method embodiments described above and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0109] Figure 4 To illustrate the hardware structure of the electronic device according to the embodiments of this application, the electronic device includes: The processor 401 can be implemented using a general-purpose central processing unit (CPU), microprocessor, application specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 402 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 402 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 402 and called by the processor 401 to execute the protein sequence screening method of the embodiments of this application. Input / output interface 403 is used to implement information input and output; The communication interface 404 is used to enable communication and interaction between this electronic device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 405 transmits information between various components of an electronic device (e.g., processor 401, memory 402, input / output interface 403, and communication interface 404); The processor 401, memory 402, input / output interface 403 and communication interface 404 are connected to each other within the electronic device via bus 405.

[0110] The electronic device provided in this application embodiment can implement each step of the protein sequence screening method embodiment described above and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0111] This application also provides a storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various steps of the protein sequence screening method embodiments described above and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0112] The processor is the processor in the electronic device described in the above embodiments. The storage medium may include a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.

[0113] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various steps of the protein sequence screening method embodiment described above, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0114] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0115] This application provides a computer program product stored in a storage medium. The program product is executed by at least one processor to implement the various steps of the protein sequence screening method embodiment described above, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0116] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not delete other identical elements present in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0117] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.

[0118] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for screening protein sequences, characterized in that, include: In response to receiving a sequence screening task for a target protein, the screening evidence package and candidate sequences of the target protein are obtained; The screening evidence package includes evidence data for characterizing the features of the target protein, and a first mapping relationship between the residue positions of the target protein and the source of the evidence data, and the residue positions of the candidate sequence; Based on the filtering evidence package, a first filtering rule is generated and output; the first filtering rule is used to indicate the filtering constraints that the candidate sequence needs to satisfy. In response to receiving a screening start command, the candidate sequences are screened based on the second screening rule to determine and output the sequence screening results of the target protein; The second filtering rule is determined based on the first filtering rule.

2. The screening method as described in claim 1, characterized in that, The screening evidence package includes multiple sets of evidence data; each set of evidence data comes from a different source; the first mapping relationship includes a second mapping relationship between the target protein and the residue positions corresponding to the multiple sets of evidence data; The step of generating and outputting the first filtering rule based on the filtering evidence package includes: Obtain multiple sets of residue positions of the target protein; each set of residue positions includes multiple residue positions of the target protein that have the second mapping relationship with the residue positions corresponding to the multiple sets of evidence data; For each group of residue positions, target evidence data is determined based on the sub-evidence data at the corresponding residue positions from the multiple groups of evidence data; Based on the target evidence data of each group of residue positions and the first mapping relationship, the first screening rule is generated and output.

3. The screening method as described in claim 2, characterized in that, The determination of target evidence data based on sub-evidence data at corresponding residue positions from the multiple sets of evidence data includes at least one of the following: Based on the ranking of the sub-evidence data at the corresponding residue positions of the multiple sets of evidence data, the sub-evidence data are fused to obtain the target evidence data; In response to the detection of a selection operation of a preset screening strategy, sub-evidence data matching the preset screening strategy is selected from the sub-evidence data at the corresponding residue positions of the multiple sets of evidence data to obtain the target evidence data; the preset screening strategy includes at least one of a conservative screening strategy, an exploratory screening strategy, an engineering stability-priority screening strategy, and an activity retention-priority screening strategy.

4. The screening method as described in claim 1, characterized in that, The evidence data includes target evidence data for at least some residue positions of the target protein; The step of generating and outputting the first filtering rule based on the filtering evidence package includes: For at least some residue positions of the candidate sequence, based on the first mapping relationship and the target evidence data, the screening constraints corresponding to the residue positions are determined; Based on the screening constraints and actions corresponding to at least some of the residue positions, the first screening rule is generated and output; the actions include at least one of elimination, retention, soft penalty, weighted sorting, multi-level grouping, labeling, and risk warning.

5. The screening method as described in claim 1, characterized in that, Before filtering the candidate sequences based on the second filtering rule, the method further includes: In response to receiving a third filtering rule input by the user, the first filtering rule and the third filtering rule are compiled to obtain the second filtering rule; The first and third filtering rules have the same structured fields; the compilation process includes at least one of the following: field integrity verification, residue position validity verification and mapping verification, filtering constraint expansion, strength interpretation, duplicate item merging and conflict detection, and execution order sorting.

6. The screening method as described in claim 1, characterized in that, The first screening rule includes screening rules targeting at least one scope of action in residue sites, residue segments, site tags, structural domain segments, and the entire sequence; Before filtering the candidate sequences based on the second filtering rule, the method further includes: In response to detecting a selection operation targeting the target's scope of action, a target filtering rule corresponding to the target's scope of action is determined and output based on the first filtering rule. Based on the target filtering rules, the second filtering rule is determined.

7. The screening method as described in claim 1, characterized in that, Before filtering the candidate sequences based on the second filtering rule, the method further includes: In response to detecting a modification operation against the first filtering rule, the second filtering rule is determined based on the modified first filtering rule; The modification operation includes performing at least one of the following operations on at least a portion of the rule content in the first filtering rule: adding, editing, disabling, deleting, confirming, or copying.

8. The screening method as described in claim 1, characterized in that, The sequence filtering results include at least one of the following: the retained candidate sequences, the eliminated candidate sequences, the sorting results of the candidate sequences, and the grouping results of the candidate sequences; After filtering the candidate sequences based on the second filtering rule, the method further includes: Generate and output audit information related to the sequence screening results; The audit information includes at least one of the following: the source information of the second screening rule, the modification record of the first screening rule, and the screening basis of the candidate sequence.

9. The screening method as described in claim 1, characterized in that, The obtained and filtered evidence package includes: Acquire first evidence data and second evidence data; the first evidence data is used to characterize the target protein in at least one dimension of three-dimensional structure and conformation features and intermolecular interactions; the second evidence data is used to characterize the target protein in at least one dimension of sequence, evolution, and function. Standardize the first and second evidence data; The filtered evidence package is generated based on the standardized first evidence data, the second evidence data, and the first mapping relationship.

10. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and the computer program, when executed by the processor, implements the screening method as described in any one of claims 1 to 9.

11. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the screening method as described in any one of claims 1 to 9.

Images