CN-121747702-B - Method, device, equipment and medium for predicting interaction between nano antibody and antigen

CN121747702BCN 121747702 BCN121747702 BCN 121747702BCN-121747702-B

Abstract

The application relates to a method, a device, computer equipment and a storage medium for predicting interaction between a nano antibody and an antigen. The method comprises the steps of obtaining a nanobody sequence to be detected and an antigen sequence to be detected, utilizing a first intercalation generator to respectively generate a nanobody intercalation vector and an antigen intercalation vector according to the nanobody sequence to be detected and the antigen sequence to be detected, utilizing a second intercalation generator to generate a nanobody antigen pair intercalation vector according to the nanobody sequence to be detected and the antigen sequence to be detected, fusing the nanobody intercalation vector and the nanobody antigen pair intercalation vector to obtain a nanobody fusion characteristic, fusing the antigen intercalation vector and the nanobody antigen pair intercalation vector to obtain an antigen fusion characteristic, and predicting interaction between the nanobody sequence to be detected and the antigen sequence to be detected according to the nanobody fusion characteristic and the antigen fusion characteristic. The method can improve the accuracy of the prediction of the interaction between the nano antibody and the antigen.

Inventors

ZHANG NING
DU YISHAN
SUN WENJIE
WAN WUZHOU

Assignees

云核医药(天津)有限公司
云南白药集团股份有限公司

Dates

Publication Date: 20260508
Application Date: 20260225

Claims (9)

1. A method of predicting nanobody interaction with an antigen, the method comprising: obtaining a nano antibody sequence to be detected and an antigen sequence to be detected; According to the nanometer antibody sequence to be detected and the antigen sequence to be detected, respectively generating a nanometer antibody embedding vector and an antigen embedding vector by using a first embedding generator, wherein the first embedding generator is an embedding generator of a protein language model; generating a nanobody antigen pair embedding vector by using a second embedding generator according to the nanobody sequence to be detected and the antigen sequence to be detected, wherein the second embedding generator is an embedding generator of a structure prediction model; The method comprises the steps of carrying out fusion on an embedded vector by a nanobody embedding vector and a nanobody antigen to obtain a nanobody fusion characteristic, carrying out fusion on the embedded vector by the nanobody embedding vector and the nanobody antigen to obtain an antigen fusion characteristic, wherein the embedded vector by the nanobody embedding vector and the nanobody antigen respectively comprises a first dimension parameter, a second dimension parameter and a third dimension parameter, carrying out fusion on the embedded vector by the nanobody embedding vector and the nanobody antigen to obtain a nanobody fusion characteristic, and carrying out interchange on the arrangement sequence of the second dimension parameter and the third dimension parameter of the nanobody embedding vector to obtain an adjusted nanobody embedding vector; And predicting the interaction between the nanobody sequence to be detected and the antigen sequence to be detected according to the nanobody fusion characteristics and the antigen fusion characteristics.
2. The method of claim 1, wherein the first dimension parameter represents a batch size, the second dimension parameter represents a longest length of a sequence, and the third dimension parameter represents a hidden dimension of a feature.
3. The method of claim 1, wherein said calculating antibody fusion weights from said adjusted nanobody intercalation vector and said adjusted nanobody antigen-pair intercalation vector comprises: Adding the adjusted nanobody embedded vector and the adjusted nanobody antigen to the embedded vector to obtain an added feature; Sequentially performing first convolution processing, batch normalization processing and activation function processing on the added features to obtain intermediate output data; And sequentially carrying out second convolution processing, batch normalization processing and activation function processing on the intermediate output data to obtain the antibody fusion weight.
4. The method of claim 1, wherein the antibody fusion weights comprise a first dimension parameter, a second dimension parameter, and a third dimension parameter, wherein the computing the nanobody fusion signature from the nanobody intercalation vector, the nanobody antigen pair intercalation vector, and the antibody fusion weights comprises: Exchanging the arrangement sequence of the second dimension parameter and the third dimension parameter of the antibody fusion weight to obtain an adjusted antibody fusion weight; And calculating according to the adjusted antibody fusion weight, the nanobody intercalation vector and the nanobody antigen pair intercalation vector to obtain the nanobody fusion characteristic.
5. The method of claim 1, wherein the antigen-embedded vector and the nanobody antigen-pair embedded vector each comprise a first dimension parameter, a second dimension parameter, and a third dimension parameter, and wherein the fusing the antigen-embedded vector and the nanobody antigen-pair embedded vector to obtain an antigen fusion feature comprises: Exchanging the arrangement sequence of the second dimension parameter and the third dimension parameter of the antigen embedding vector to obtain an adjusted antigen embedding vector; The arrangement sequence of the second dimension parameter and the third dimension parameter of the nanobody antigen pair embedding vector is exchanged to obtain an adjusted nanobody antigen pair embedding vector; Calculating antigen fusion weights according to the adjusted antigen embedding vectors and the adjusted nanobody antigen-pair embedding vectors; and calculating according to the antigen embedding vector, the nanobody antigen pair embedding vector and the antigen fusion weight to obtain the antigen fusion characteristic.
6. The method of claim 1, wherein said predicting interactions between said nanobody sequences under test and said antigen sequences under test based on said nanobody fusion characteristics and said antigen fusion characteristics comprises: splicing the nanobody fusion characteristic and the antigen fusion characteristic to obtain a splicing characteristic; respectively carrying out first pooling, second pooling and third pooling on the splicing characteristics; performing first sensing mechanism processing on the first pooled output, performing second sensing mechanism processing on the second pooled output, and performing third sensing mechanism processing on the third pooled output to respectively obtain a first processing result, a second processing result and a third processing result; And obtaining a predicted result of the interaction between the nanometer antibody sequence to be detected and the antigen sequence to be detected according to the first processing result, the second processing result and the third processing result.
7. A nanobody and antigen interaction prediction device, comprising: The data acquisition module is used for acquiring a nano antibody sequence to be detected and an antigen sequence to be detected; A first embedding generation module for generating nanobody embedding vectors and antigen embedding vectors respectively by using a first embedding generator according to the nanobody sequence to be detected and the antigen sequence to be detected, wherein the first embedding generator is an embedding generator of a protein language model The second embedding generation module is used for generating a nanobody antigen pair embedding vector by utilizing a second embedding generator according to the nanobody sequence to be detected and the antigen sequence to be detected, wherein the second embedding generator is an embedding generator of a structure prediction model; The characteristic fusion module is used for fusing the nanobody embedding vector and the nanobody antigen pair embedding vector to obtain a nanobody fusion characteristic, fusing the antigen embedding vector and the nanobody antigen pair embedding vector to obtain an antigen fusion characteristic, wherein the nanobody embedding vector and the nanobody antigen pair embedding vector both comprise a first dimension parameter, a second dimension parameter and a third dimension parameter, the characteristic fusion module is specifically used for exchanging the arrangement sequence of the second dimension parameter and the third dimension parameter of the nanobody embedding vector to obtain an adjusted nanobody embedding vector, exchanging the arrangement sequence of the second dimension parameter and the third dimension parameter of the nanobody antigen pair embedding vector to obtain an adjusted nanobody antigen pair embedding vector, and calculating an antibody fusion weight according to the adjusted nanobody embedding vector and the adjusted nanobody antigen pair embedding vector; And the integrated decision module is used for predicting the interaction between the nano antibody sequence to be detected and the antigen sequence to be detected according to the nano antibody fusion characteristics and the antigen fusion characteristics.
8. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any one of claims 1 to 6 when the computer program is executed by the processor.
9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.

Description

Method, device, equipment and medium for predicting interaction between nano antibody and antigen Technical Field The application relates to the technical field of biological information, in particular to a method, a device, computer equipment and a storage medium for predicting interaction between a nano antibody and an antigen. Background Nanobodies (nanobodies), also known as variable regions (VHH) of heavy chain antibodies, are a unique single domain fragment derived from antibodies that contain only heavy chains naturally occurring in, for example, camelids. They have a smaller molecular weight (12-15 k Da) and only three complementarity determining regions (CDR 1, CDR2, and CDR 3), nanobodies have significant advantages over conventional antibodies in terms of tissue penetration, blood brain barrier permeability, stability, and production costs. In recent years, nanobodies have been attracting attention due to their wide application in the biomedical field, which application encompasses sensitive detection methods and novel therapeutic strategies. With the continued penetration of nanobody research, the continued release of related public databases has driven advances in algorithmic research. In the prediction of nanobody-antigen interactions, computational methods undergo a paradigm shift from traditional methods to machine learning and deep learning models. However, in the prior art, based on machine learning and deep learning models, prediction is performed only by relying on energy scores or sequence information, but these methods fail to fully utilize multi-modal data, resulting in limited prediction capability and less than expected prediction accuracy of the models. Disclosure of Invention In view of the foregoing, it is desirable to provide a nanobody-antigen interaction prediction method, apparatus, computer device, and storage medium that can improve the accuracy of nanobody-antigen interaction prediction. In a first aspect, there is provided a method of predicting nanobody interaction with an antigen, the method comprising: obtaining a nano antibody sequence to be detected and an antigen sequence to be detected; According to the sequence of the nano antibody to be detected and the sequence of the antigen to be detected, respectively generating a nano antibody embedding vector and an antigen embedding vector by using a first embedding generator, wherein the first embedding generator is a protein embedding generator; generating a nanobody antigen pair embedding vector by using a second embedding generator according to the nanobody sequence to be detected and the antigen sequence to be detected, wherein the second embedding generator is a structure prediction embedding generator; fusing the nanobody intercalation vector and the nanobody antigen-pair intercalation vector to obtain a nanobody fusion characteristic, and fusing the antigen intercalation vector and the nanobody antigen-pair intercalation vector to obtain an antigen fusion characteristic; And predicting the interaction between the nanometer antibody sequence to be detected and the antigen sequence to be detected according to the nanometer antibody fusion characteristics and the antigen fusion characteristics. In some embodiments, the nanobody intercalation vector and the nanobody antigen pair intercalation vector each include a first dimension parameter, a second dimension parameter, and a third dimension parameter, and the fusion of the nanobody intercalation vector and the nanobody antigen pair intercalation vector results in nanobody fusion characteristics comprising: The arrangement sequence of the second dimension parameter and the third dimension parameter of the nanobody embedding vector is exchanged to obtain an adjusted nanobody embedding vector; the arrangement sequence of the second dimension parameter and the third dimension parameter of the nanobody antigen pair embedding vector is exchanged to obtain an adjusted nanobody antigen pair embedding vector; calculating antibody fusion weights according to the adjusted nanobody intercalation vectors and the adjusted nanobody antigen-to-intercalation vectors; And calculating according to the nanobody intercalation vector, the nanobody antigen pair intercalation vector and the antibody fusion weight to obtain the nanobody fusion characteristic. In some embodiments, the first dimension parameter represents a lot size, the second dimension parameter represents a longest length of the sequence, and the third dimension parameter represents a hidden dimension of the feature. In some embodiments, calculating antibody fusion weights from the adjusted nanobody intercalation vector and the adjusted nanobody antigen-to-intercalation vector comprises: Adding the adjusted nanobody embedded vector and the adjusted nanobody antigen to the embedded vector to obtain an added characteristic; sequentially performing first convolution processing, batch normalization processing and activation func