22 results about "Biocomputer" patented technology

The present disclosure provides a polypeptide generation method, device, electronic equipment and storage medium, relates to the technical field of artificial intelligence, in particular to biological computing technology. The method comprises the following steps: obtaining a protein target and a reference polypeptide corresponding to the protein target, wherein the protein target is a protein causing pathological changes; generating a candidate polypeptide corresponding to the reference polypeptide; and inputting the protein target and the candidate polypeptide into an affinity evaluation module for screening to obtain a target polypeptide, wherein the affinity evaluation module comprises at least two levels of evaluation sub-modules arranged in a preset evaluation order, and the later the evaluation sub-module in the evaluation order, the higher the accuracy of the evaluation sub-module. The present disclosure can balance the accuracy and efficiency, and improve the "performance-price ratio" of the polypeptide drug design process.

The application discloses a DNA nanocage probe for leukemia cell multi-specific membrane protein logic circuit analysis, a kit and a method thereof, and realizes in-situ imaging analysis on three target membrane proteins by using a probe of a cross-section octahedral DNA nanocage main frame, stable structure, and designing a three-specific recognition biological computing DNA molecule logic gate based on a nucleic acid aptamer (Sgc4f, TC01, Sgc8c), which provides a new tool and idea for early diagnosis of tumors and biomedical application of the DNA molecule logic gate in a complex cell system.

The application discloses a dopamine transporter dynamics mechanism-based mental stimulant addiction risk prediction method, and belongs to the field of computer-aided drug design and biological computing simulation. The method comprises the following steps: firstly, preparing a conformation and constructing a drug-target-membrane environment; then, performing classical molecular dynamics (cMD) simulation and binding affinity evaluation; then, extracting a dissociation energy barrier by using stretching molecular dynamics (SMD); then, extracting a binding path based on a ligand binding-parallel cascade molecular dynamics (LB-PaCS-MD) and a Markov state model; and finally, comprehensively evaluating addiction and screening drugs. By using the method, a dynamic dissociation / binding path can be accurately mapped on a full-atom level, and a residence time parameter can be calculated, so as to guide accurate development of low-addiction mental drugs.

The present disclosure provides a molecule generation method and device, electronic equipment and storage medium, and relates to the fields of artificial intelligence such as biological computing and deep learning. The method can include: obtaining a first molecule to be processed, and obtaining a synthesis route of the first molecule as an original synthesis route; modifying the original synthesis route to obtain a target synthesis route; and generating a second molecule required according to the target synthesis route. The scheme disclosed in the present disclosure can improve the generation efficiency of molecules and the quality of generated molecules.

The application belongs to the technical field of biological computing, and discloses a ternary addition system and a computing method based on DNA strand displacement, which comprises an input unit, a computing unit and an output unit. The input unit adds the input information to the corresponding computing module. The computing unit comprises nine modules, each of which is provided with a competitive blocking circuit and a carry computing gate, and dynamically selects a result bit 1 or 2 and generates a carry signal. The output unit outputs the digital information of 0-2 according to the fluorescent signal of the result bit, extracts the next bit carry information generated by the carry computation, amplifies the signal and then adds it to the next bit computation. The application breaks through the limitation of the number of DNA computing bits through the cooperation of the ternary gate control architecture and the fuel chain amplification, can accurately calculate the ternary addition, can expand the number of bits calculated, can realize 10-bit addition operation, and can be integrated with the expansion of the multiplication logic, thereby providing a universal framework for the molecular arithmetic unit.

The application discloses a complex structure prediction method and device and a model training method and an electronic equipment, relates to the technical field of artificial intelligence, in particular to the technical field of deep learning and biological computing. The specific implementation scheme is as follows: obtaining the amino acid sequence of a receptor and the sequence information of a ligand to be combined, and the protein structure of the receptor, determining the orientation information of the ligand relative to the protein structure based on the protein structure of the receptor, and determining the target complex structure of the combination of the receptor and the ligand according to the orientation information of the ligand relative to the protein structure, the amino acid sequence of the receptor and the sequence information of the ligand.

The invention belongs to the technical field of biological computer analysis, and discloses an analysis method based on network pharmacology and a molecular docking technology. The psoriasis treatment of fucoidin is analyzed on the basis of network pharmacology and a molecular docking technology, five core targets, 30 GO annotations and 66 KEGG pathways are screened through the network pharmacology and the molecular docking technology, and the affinity of active ingredients and the core targets is good.

A model-based peptide design method in the field of artificial intelligence technology such as biological computing is provided. The specific implementation includes: obtaining a pocket of an objective target protein and an objective peptide, a reserved position for designing a unnatural amino acid is identified in the objective peptide, and the pocket binds to the objective peptide via the reserved position; obtaining a feature of the pocket of the objective target protein and multimodal features of each known amino acid in the objective peptide; the multimodal features of each known amino acid comprise a backbone orientation feature, a backbone rotation feature, a side chain type feature, and a rigid atom group distribution feature; designing the unnatural amino acid at the reserved position in the objective peptide using a pre-trained peptide design model based on the feature of the pocket of the objective target protein and the multimodal features of each known amino acid in the objective peptide.

The present disclosure provides a model-based polypeptide design method, model training method, device and equipment, and particularly relates to the field of artificial intelligence technologies such as biological computing. The specific implementation scheme is: obtaining a pocket of a target target protein and a target polypeptide, the target polypeptide being marked with a reserved position of a non-natural amino acid to be designed; the pocket is combined with the target polypeptide through the reserved position; obtaining the characteristics of the pocket of the target target protein and the multi-modal characteristics of each known amino acid in the target polypeptide; the multi-modal characteristics of each known amino acid include backbone direction characteristics, backbone rotation characteristics, side chain type characteristics, and rigid atom group distribution characteristics; based on the characteristics of the pocket of the target target protein and the multi-modal characteristics of each known amino acid in the target polypeptide, a pre-trained polypeptide design model is used to design a non-natural amino acid at the reserved position in the target polypeptide. The technology of the present disclosure can effectively improve the accuracy of the non-natural amino acid at the reserved position in the designed target polypeptide.

This invention relates to synthetic biology, gene and cell therapy, and particularly to a gene translation system regulated by non-natural amino acids and its applications. This gene translation system precisely regulates gene expression initiation using non-natural amino acids. The invention also provides eukaryotic expression vectors, mammalian cell lines, microcapsules, vacuum fiber tubes, and non-natural amino acid cookies containing the aforementioned gene translation system. Furthermore, this invention discloses that the multifunctional platform can precisely regulate insulin expression and release for diabetes treatment; this gene expression regulation platform can be used to construct complex biological computer-controlled systems. This invention provides a powerful new gene expression regulation tool for gene therapy and cell therapy.

The application belongs to the field of biological computer technology and DNA nanotechnology, and particularly relates to a logic element based on a DNA nanostructure and application of the logic element in manufacturing a true random number generator. The logic element based on the DNA nanostructure takes a two-dimensional cross DNA nanostructure as a main body of the element, utilizes basic characteristics of the DNA nanostructure, such as uniform size, accurate site programming and rich extensible edge chains, constructs logic elements carrying different information through site programming, takes the extensible edge chains as sticky ends to realize a stable and reliable coaxial physical connection relationship between the logic elements, and only when there is a complementary relationship between the sticky ends of the logic elements, can a dimer or a multimer be formed through self-assembly. Therefore, when the logic element is applied to manufacturing the random number generator, the multimer connected by the cross paper folds still has good rigidity, and the correctness of the random generation result is ensured.

The application discloses a biological computing multi-tool unified calling method and system based on a large language model, relates to the fields of biological information and computational chemistry, and at least comprises a task planner, an executor and a tool layer three parts, the method applied to the tool layer in the biological computing multi-tool unified calling system based on the large language model comprises the following steps: based on a unified interactive interface, responding to an interactive instruction with the executor, calling a corresponding type tool to return an execution result of a to-be-executed task step to the executor, wherein, when executing each to-be-executed task in a sequence of executable steps step by step, the executor triggers an interactive instruction with the tool layer based on the type of each to-be-executed task step, and the sequence of executable steps is obtained after the task planner sequentially performs semantic analysis, determines corresponding targets and constraints and performs corresponding task decomposition processing on a to-be-processed file carried in a user input instruction. The application improves task processing efficiency.

The application discloses a training method and device of a protein structure generation model, a protein structure generation method and device, and relates to the technical field of computers, in particular to the fields of artificial intelligence such as biological computing and deep learning. The specific implementation scheme is as follows: a noisy protein structure and a reference semantic representation of an original protein structure corresponding to the noisy protein structure are obtained; in the process of denoising the noisy protein structure by using an initial protein structure generation model, an intermediate hidden variable representation is extracted from the initial protein structure generation model; and the initial protein structure generation model is trained according to the intermediate hidden variable representation and the reference semantic representation, so as to obtain a trained protein structure generation model.

The present disclosure provides a protein design method, device, equipment and medium, relates to the field of artificial intelligence, in particular to the technical field of deep learning, biological computing and large language model. The generation method comprises the following steps: constructing a plurality of candidate proteins, each of which comprises a first chain of an original protein and a non-natural sequence constructed based on a second chain of the original protein; retrieving a first multiple sequence alignment of the first chain and a second multiple sequence alignment of the non-natural sequence; matching the first multiple sequence alignment and the second multiple sequence alignment to obtain a cross-chain homologous sequence by using a pre-trained initial protein language model; predicting the structure and a first score of the candidate protein by using a protein structure prediction model; determining a reward value based on the first score and performing reinforcement learning training on the initial protein language model; determining a second score of each of the plurality of candidate proteins by using the trained target protein language model and the protein structure prediction model, so as to obtain a protein design result.

The invention discloses a method for predicting the harmfulness of synonymous mutation based on multilevel biological characteristics. The method comprises the following steps: 1, collecting and preprocessing a synonymous mutation sample; 2, carrying out DNA, RNA and protein level feature annotation on the obtained synonymous mutation sample by using a biological calculation tool, and carrying out pretreatment; 3, performing dimension reduction on the high-dimensional biological characteristic data by using a causal characteristic selection algorithm to obtain a plurality of causal characteristic subsets; 4, splicing an optimal feature set and performing model construction and training by using a machine learning classifier; and 5, predicting an external test set sample by using the trained optimal model to obtain a synonymous mutation harmfulness prediction score. According to the method, the features having the causal relationship with the target variable can be screened out from the multi-level high-dimensional biological features, and then the causal features of the three biological levels are fused for prediction, so that the synonymous mutation harmfulness prediction precision is improved.

This disclosure provides a training method, design method, apparatus, and device for a protein reverse folding model, relating to the field of computer science, and particularly to artificial intelligence and biological computing technologies. The training method includes: acquiring a training sample set, which includes structural information of the target protein and the binding protein; inputting the training samples into the protein reverse folding model to obtain multiple candidate amino acid sequences of the binding protein; calculating the binding performance indices of the target protein and each candidate amino acid sequence of the binding protein, including complex conformation quantification indices and / or binding affinity quantification indices; constructing positive and negative sample pairs based on the binding performance indices; and training the protein reverse folding model based on a reinforcement learning algorithm and the positive and negative sample pairs to obtain the trained protein reverse folding model. This disclosure can improve the ability of protein reverse folding models to generate binding protein sequences with superior binding performance.

The application discloses a kind of RNA small molecule binding site prediction method based on multiscale geometric deep learning, belong to bioinformatics and structural biological computing field.The method obtains RNA-small molecule complex structure data and constructs residue level binding site label, extracts RNA sequence information and atomic three-dimensional structure features, generates sequence representation using pre-training RNA language model, constructs deep learning model including hybrid atom encoder, attention type atom to residue polymerization module, residue level space graph neural network module and geometric bias global attention module, fuses atomic level, residue level and global multiscale information, carries out binding site prediction to RNA residue, outputs binding probability and prediction uncertainty.The method can improve prediction accuracy and generalization ability, and is suitable for RNA target drug design and structure-guided small molecule screening.

This invention discloses a molecular representation method and system based on deep graph neural networks, relating to the field of artificial intelligence technology, particularly to the fields of biological computing and deep learning. It addresses the problem in existing technologies where excessive smoothing prevents graph neural networks from increasing the number of graph convolutions to expand the receptive field of nodes. This invention utilizes a third-party library to convert computer-stored molecular data into a molecular graph, which includes an adjacency matrix, nodes, and edges. The molecular graph is analyzed to obtain initial feature vectors for the nodes and edges. Based on the graph neural network and the adjacency matrix, dense residual graph convolution operations are performed on the initial feature vectors of the nodes and edges to obtain new feature vectors for the nodes and edges. Pooling operations are then performed on the new feature vectors of the nodes and edges to obtain the molecular representation. This invention is used for molecular representation.

The invention discloses a biological computing multi-tool unified calling method and system based on a large language model, and relates to the field of biological information and computational chemistry, and the system at least comprises a task planner, an actuator and a tool layer. The method applied to the tool layer in the biological computing multi-tool unified calling system based on the large language model comprises the steps that based on a unified interaction interface, in response to an interaction instruction with an actuator, a tool of a corresponding type is called to return an execution result of a task step to be executed to the actuator, when the executor gradually executes each step of to-be-executed task in the executable step sequence, an interaction instruction with the tool layer is triggered based on the step type of each step of to-be-executed task; the executable step sequence is obtained after the task planner performs semantic analysis, corresponding target and constraint determination and corresponding task decomposition processing on a to-be-processed file carried in a user input instruction in sequence. The task processing efficiency is improved.

The invention belongs to the technical field of biological computing, and discloses a general graph problem solving method and system based on DNA strand displacement, and the method comprises the steps: S1, carrying out graph structure modeling, and carrying out graph representation module based on DNA strand displacement; s2, constructing a detection module, and realizing parallel biochemical identification of constraint conditions by using a detection gate D (u) containing a fluorescence label; and S3, carrying out parallel detection on candidate solutions, supporting a unified solution framework of problems such as a minimum control set and a maximum independent set, and realizing rapid screening of the solutions through candidate solution enumeration and a fluorescence signal criterion. The method and the system provided by the invention adopt a modular cascade structure, realize direct solution of a complex problem on the premise of not depending on electronic calculation conversion, compress a solution space in combination with a graph feature value preprocessing technology, have the technical advantages of high neighborhood recognition accuracy, strong leakage reaction inhibition, strong experimental process universality and the like, and have a wide application prospect. And the application efficiency and stability of DNA calculation in the field of combinatorial optimization are remarkably improved.

The present disclosure provides a binding conformation prediction method, a model training method, an apparatus and a storage medium, which relate to the field of computer technology, in particular to the field of artificial intelligence and biological computing. The specific implementation scheme is: based on the binding conformation sample set generated by the target protein, the parameters of the first binding conformation prediction model are adjusted to obtain a second binding conformation prediction model; a plurality of candidate binding conformations corresponding to the target protein are respectively input into the second binding conformation prediction model to obtain a prediction result for each candidate binding conformation; based on the prediction result of each candidate binding conformation, the target binding conformation of the target protein is obtained. According to the embodiment of the present disclosure, the first binding conformation prediction model can be adjusted based on the binding conformation sample set generated by the target protein to obtain a second binding conformation prediction model, thereby improving the accuracy of the second binding conformation prediction model's prediction result for the binding conformation of the target protein.