WO2024250183A1 - 一种基于图神经网络的药物抑癌敏感性预测方法与装置 - Google Patents

一种基于图神经网络的药物抑癌敏感性预测方法与装置 Download PDF

Info

Publication number
WO2024250183A1
WO2024250183A1 PCT/CN2023/098718 CN2023098718W WO2024250183A1 WO 2024250183 A1 WO2024250183 A1 WO 2024250183A1 CN 2023098718 W CN2023098718 W CN 2023098718W WO 2024250183 A1 WO2024250183 A1 WO 2024250183A1
Authority
WO
WIPO (PCT)
Prior art keywords
drug
cancer cell
cell line
feature
neural network
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2023/098718
Other languages
English (en)
French (fr)
Inventor
郑鹏飞
劳传祺
陈红阳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang Lab
Original Assignee
Zhejiang Lab
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhejiang Lab filed Critical Zhejiang Lab
Priority to PCT/CN2023/098718 priority Critical patent/WO2024250183A1/zh
Publication of WO2024250183A1 publication Critical patent/WO2024250183A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/30—Prediction of properties of chemical compounds, compositions or mixtures

Definitions

  • the present invention mainly relates to the field of artificial intelligence drug research and development, and specifically to a method and device for predicting drug cancer inhibition sensitivity based on graph neural network.
  • Drug anticancer sensitivity prediction based on deep learning is currently a hot topic. This type of method inputs drug and cancer cell line data into a deep learning model to form a "drug-cancer cell line" instance pair, and uses a deep learning algorithm to extract features and mine the internal connection of the instance pair to predict drug anticancer sensitivity.
  • the current deep learning method for drug anticancer sensitivity has the following limitations:
  • Cancer cell lines are difficult to characterize, and being represented by only a small amount of cancer cell line omics data will lead to insufficient feature extraction of cancer cell lines.
  • the present invention proposes a method and device for predicting drug cancer inhibition sensitivity based on graph neural network.
  • a method for predicting drug cancer inhibition sensitivity based on a graph neural network comprising:
  • a graph neural network model is constructed based on the drug adjacency matrix, node feature matrix and edge feature matrix, and the drug graph features are obtained by iterating the graph neural network model.
  • the drug graph features are aggregated with the drug molecule feature vector to obtain the drug representation.
  • the drug-cancer cell line example pairs are input into the drug tumor suppressor sensitivity prediction model to predict the IC50 value of the candidate drug in the cancer cell line environment.
  • a device for predicting drug cancer inhibition sensitivity based on a graph neural network comprising: a multi-source data integration module, a storage module and a processing module;
  • the multi-source data integration module is used to obtain the original data of candidate drugs and cancer cell lines; extract the drug information features of candidate drugs and expand them; construct the drug molecule feature vector based on the drug information features, and construct the drug adjacency matrix, node feature matrix and edge feature matrix with the drug atoms as nodes and chemical bonds as edges;
  • the storage module includes: a drug information database and a cancer cell line information database;
  • Drug information database used to store drug molecule feature vectors, drug adjacency matrix, node feature matrix and edge feature matrix;
  • Cancer cell line information database used to store gene expression, gene mutation, gene methylation, and gene copy number expression of gene sequences in cancer cell lines;
  • the processing modules include: a drug characterization extraction unit, a cancer cell line characterization extraction unit, a drug-cancer cell line instance pair acquisition unit, and a drug tumor suppressor sensitivity prediction unit;
  • Drug characterization extraction unit construct a graph neural network model based on the drug adjacency matrix, node feature matrix and edge feature matrix, and iterate the graph neural network model to obtain drug graph features; after the drug graph features are aggregated with the drug molecule feature vector, the drug characterization is obtained;
  • a cancer cell line characterization extraction unit extracts cancer cell line characterization by obtaining the expression of gene expression, gene mutation, gene methylation, and gene copy number of gene sequences in cancer cell lines;
  • a drug-cancer cell line instance pair acquisition unit used for performing feature fusion on drug representation and cancer cell line representation to obtain a drug-cancer cell line instance pair;
  • the drug tumor suppressor sensitivity prediction unit predicts the IC50 value of the candidate drug in the cancer cell line environment by inputting the drug-cancer cell line instance pair into the drug tumor suppressor sensitivity prediction model.
  • an electronic device comprising one or more processors, for implementing the above-mentioned method for predicting drug cancer inhibition sensitivity based on graph neural network.
  • a computer-readable storage medium on which a program is stored.
  • the program is executed by a processor, it is used to implement the above-mentioned drug cancer inhibition sensitivity prediction method based on graph neural network.
  • the present invention provides a method and device for predicting drug tumor suppressor sensitivity based on graph neural network, which can extract drug information of candidate drugs in view of the difficulty in obtaining drug data, incomplete drug feature information, and sparse drug features.
  • Information features, and expanded, integrated major drug information data sources breaking the bottlenecks of few drug information features in a single data source, missing some drug information features, etc., fully capturing drug feature information, and increasing the dimension of drug information features;
  • the expression of gene expression, gene mutation, gene methylation, and gene copy number of gene sequences in cancer cell lines is used to accurately represent cancer cell lines, and it is encoded through Transformer network and feedforward neural network to extract cancer cell line characterization.
  • the present invention performs feature fusion on drug characterization and cancer cell line characterization to obtain drug-cancer cell line example pairs, and significantly improves the prediction accuracy of drug anti-cancer sensitivity through more accurate and richer drug-cancer cell line example pairs.
  • FIG1 is a flow chart of a method for predicting drug cancer inhibition sensitivity based on a graph neural network provided by an embodiment of the present invention
  • FIG2 is a flowchart of obtaining drug characterization according to an embodiment of the present invention.
  • FIG3 is a flowchart of obtaining cancer cell line characterization according to an embodiment of the present invention.
  • FIG4 is a schematic diagram of a drug cancer inhibition sensitivity prediction device based on a graph neural network provided by an embodiment of the present invention.
  • FIG. 5 is a schematic diagram of an electronic device provided by an embodiment of the present invention.
  • the embodiment provides a method and device for predicting drug tumor suppressor sensitivity based on graph neural network, which integrates major drug information data sources, breaks the bottlenecks of few drug information features and missing drug information features of a single data source, fully captures drug feature information, and increases the dimension of drug information features; accurately represents cancer cell lines through the expression of gene expression, gene mutation, gene methylation, and gene copy number of gene sequences in cancer cell lines, and encodes them through Transformer and feedforward neural networks.
  • the present invention significantly improves the prediction results of drug tumor suppressor sensitivity by extracting more accurate and richer "drug-cancer cell line" representations.
  • FIG1 shows a flow chart of a method for predicting drug cancer inhibition sensitivity based on a graph neural network provided by an embodiment of the present invention.
  • the method specifically comprises the following steps:
  • Step S1 obtaining original data of candidate drugs and cancer cell lines.
  • Step S2 extracting drug information features of candidate drugs and expanding them; constructing drug molecular feature vectors based on drug information features, taking drug atoms as nodes and chemical bonds as edges, constructing drug adjacency matrix, node feature matrix and edge feature matrix.
  • step S2 includes the following sub-steps:
  • Step S201 using the SMILES sequence of the molecule as the input standard, obtain drug information features from various drug information data sources (DeepChem, RDKit, PubChem, etc.), and clarify the specific meaning of each dimension in the drug information features by reading the official documents of each drug information data source.
  • drug information data sources EdepChem, RDKit, PubChem, etc.
  • the extractable drug information features include the atomic characteristics of the drug (atom type, formal charge, hybrid orbital, etc.), the chemical bond characteristics of the drug (bond type, homocyclicity, conjugation, etc.), and the molecular characteristics of the drug (molecular weight, number of valence electrons, etc.).
  • the feature dimension is 30 dimensions, of which dimensions 1-10 represent the atom type, dimension 11 represents the formal charge number, dimensions 12-14 represent hybrid orbitals, dimensions 15-16 represent hydrogen bond donors or acceptors, dimension 17 indicates whether the atom belongs to an aromatic ring, dimensions 18-24 represent the degree of the atom, and dimensions 25-30 represent the number of hydrogen atoms connected to the atom.
  • Step S202 write a standard document of extractable drug information features, and remove duplicates, complete, and merge the drug information features obtained in step S201.
  • Step S203 based on the drug information obtained in step S202, with the atoms of the drug as nodes and the chemical bonds as edges, construct a drug adjacency matrix A, a node feature matrix V, an edge feature matrix E and a drug molecule feature vector Molecule_Feature.
  • the drug adjacency matrix A is given in the form of a matrix with dimension (shape, shape), where "shape” represents the number of nodes of the drug (ie, the number of atoms of the drug).
  • the node feature matrix V is given in the form of a matrix with dimensions (shape, dim1), where "dim1" represents the number of features of each node of the drug (ie, the number of features of each atom of the drug).
  • the edge feature matrix E is given in the form of an adjacency matrix with dimensions (shape, shape, dim2), where "dim2" represents the number of features of each edge of the drug (i.e., the number of features of the drug's chemical bonds). Furthermore, when there is no edge between two nodes, the corresponding coordinate elements in the edge feature matrix are filled with "0".
  • the drug molecule feature vector Molecule_Feature is given by a feature vector of dimension (1, dim3), where "Feature 3" represents the molecular feature dimension of the drug.
  • the construction of the node feature matrix V and the edge feature matrix E since the construction methods of the two are the same, the construction of the node feature matrix is taken as an example to explain.
  • this example only takes two data sources, and expresses the node feature matrices V A and V B that can be obtained by data source A and data source B in the following forms:
  • feature F B1 and F A1 in data source A are representations of the same property
  • the feature that represents more in F A1 and F B1 is used as the representation of the property.
  • the "Hybridization" attribute of atoms is only represented by the MolGraphConvFeaturizer extractor as “sp”, “sp2”, “sp3”, while the DMPNNFeaturizer extractor represents it as “sp”, “sp2”, “sp3”, “sp3d”, “sp3d2".
  • this example uses the attribute extracted from the DMPNNFeaturizer extractor as the final attribute; when features F A1 and F B1 are non-discretized data, this example directly connects F A1 and F B1 as the representation of the property, which can be expressed in the following form:
  • different molecular features can be obtained from different data sources, and these molecular features can be encoded into the drug molecule feature vector Molecule_Feature through encoding methods such as "one-hot”.
  • Step S204 the drug adjacency matrix A, the node feature matrix V, the edge feature matrix E and the drug molecule feature vector Molecule_Feature are transferred to the drug information database for storage.
  • step S3 builds a graph neural network model based on the drug adjacency matrix, node feature matrix and edge feature matrix, and iterates the graph neural network model to obtain drug graph features; after the drug graph features are aggregated with the drug molecule feature vectors, the drug representation is obtained.
  • the material information features and the corresponding drug adjacency matrix A, node feature matrix V, edge feature matrix E, etc. can be obtained from the drug information database.
  • the graph neural network model is built and iterated, and message transmission and aggregation are performed in the form of node update and edge update to obtain the drug graph feature (Graph_Feature).
  • the final drug representation is obtained.
  • step S3 includes the following sub-steps:
  • Step S301 to build and iterate the graph neural network model, perform message transmission and aggregation in the form of node update and edge update, and obtain the drug graph feature (Graph_Feature).
  • the graph neural network model is updated in the form of node update, and the node features are updated in the lth layer of the graph neural network model.
  • the expression is as follows:
  • the edge feature matrix is a matrix of dimension (shape, shape, dim2), where "dim2" represents the number of features of each edge of the drug, that is, the number of features of the drug chemical bonds; during the node update process, the edge feature matrix is split into "dim2" edge feature matrices of (shape, shape, 1);
  • V (l+1) represents the node feature matrix of the l+1th layer, Updata represents the update function, V (l) represents the node feature submatrix of the lth layer, ⁇ is the activation function, represents the ith edge feature submatrix after the edge feature matrix on the lth layer is split, Wi (l+1) and Represents the graph neural network model parameters.
  • ⁇ is the activation function
  • Vi (l+1) represents the feature vector of node i in the l+1th layer
  • Vj (l+1) represents the feature vector of node j in the l+1th layer
  • Updata represents the update function
  • Step S303 calculate the drug graph feature (Graph_Feature), the expression is as follows:
  • represents the number of nodes
  • Vi (L) represents the feature vector of node i in the Lth layer
  • L is the number of layers of the graph neural network model.
  • Whole_Feature represents drug characterization
  • Graph_Feature represents drug graph features
  • Molecule_Feature represents drug molecular features
  • step S4 is to obtain the expression levels of gene expression, gene mutation, gene methylation, and gene copy number of the gene sequence in the cancer cell line, and extract the cancer cell line characterization.
  • step S4 includes the following sub-steps:
  • Step S401 obtaining information related to the expression levels of each gene sequence in the cancer cell line.
  • its expression sequence X c1 can be expressed in the following form:
  • n represents the number of pathogenic genes in the cancer cell line
  • X c1,gi represents the expression level of gene expression (express c1,gi ), gene mutation (mutat c1,gi ), gene methylation (methy c1,gi ), and gene copy level copy c1,gi of the i-th pathogenic gene in the cancer cell line.
  • Step S402 input the expression sequence Xc1 of the cancer cell line C1 into the Transformer network, and output the representation vector Zc1 , which is expressed as follows:
  • this example uses a multi-head attention mechanism in the Transformer network.
  • Softmax is the activation function. They are the Q, K, and V matrices of the Transformer network, Zc1 is the output representation of Xc1 after passing through the Transformer network.
  • W and b represent the parameters of the feedforward neural network model
  • Mol_Cell_Feature represents the drug-cancer cell line instance pair
  • Whole_Feature represents the drug characterization
  • Cell_Feature represents the cancer cell line characterization
  • Step S6 inputting the drug-cancer cell line example pair into the drug tumor suppressor sensitivity prediction model to predict the IC50 value of the candidate drug in the cancer cell line environment.
  • the drug tumor suppressor sensitivity prediction model is a classifier; the drug-cancer cell line instance pair is input into the classifier to predict the drug tumor suppressor sensitivity, and the IC50 value of the drug in the cancer cell line environment is predicted, thereby quantifying the inhibitory ability of the drug on the cancer cell line.
  • Mol_Cell_Feature represents the drug-cancer cell line instance pair
  • W and b are the parameters of the drug tumor suppressor sensitivity prediction model.
  • the present invention provides a method for predicting drug tumor suppressor sensitivity based on graph neural network, which extracts drug information features of candidate drugs and expands them in view of the trouble of obtaining drug data, incomplete drug feature information, and sparse drug features, and integrates major drug information data sources, breaking the bottlenecks such as few drug information features of a single data source and missing drug information features of some drugs, fully capturing drug feature information, and increasing the dimension of drug information features; in view of the difficulty in characterizing cancer cell lines, the expression of gene expression, gene mutation, gene methylation, and gene copy number of gene sequences in cancer cell lines is used to accurately represent cancer cell lines, and it is encoded by Transformer network and feedforward neural network to extract cancer cell line characterization.
  • the present invention performs feature fusion on drug characterization and cancer cell line characterization to obtain drug-cancer cell line example pairs, and significantly improves the prediction accuracy of drug tumor suppressor sensitivity through more accurate and richer drug-cancer cell line example pairs.
  • the embodiment of the present invention also provides a drug cancer suppression sensitivity prediction device based on graph neural network, as shown in FIG4 , the device includes: a multi-source data integration module, a storage module and a processing module;
  • the multi-source data integration module is used to obtain the original data of candidate drugs and cancer cell lines; extract the drug information features of candidate drugs and expand them; construct the drug molecule feature vector based on the drug information features, and construct the drug adjacency matrix, node feature matrix and edge feature matrix with the drug atoms as nodes and chemical bonds as edges;
  • the storage module includes: a drug information database and a cancer cell line information database;
  • Drug information database used to store drug molecule feature vectors, drug adjacency matrix, node feature matrix and edge feature matrix;
  • Cancer cell line information database used to store gene sequences in cancer cell lines, gene expression, gene mutation, gene methylation chemistry, expression of gene copy number;
  • the processing modules include: a drug characterization extraction unit, a cancer cell line characterization extraction unit, a drug-cancer cell line instance pair acquisition unit, and a drug tumor suppressor sensitivity prediction unit;
  • Drug characterization extraction unit construct a graph neural network model based on the drug adjacency matrix, node feature matrix and edge feature matrix, and iterate the graph neural network model to obtain drug graph features; after the drug graph features are aggregated with the drug molecule feature vector, the drug characterization is obtained;
  • a cancer cell line characterization extraction unit extracts cancer cell line characterization by obtaining the expression of gene expression, gene mutation, gene methylation, and gene copy number of gene sequences in cancer cell lines;
  • a drug-cancer cell line instance pair acquisition unit used for performing feature fusion on drug representation and cancer cell line representation to obtain a drug-cancer cell line instance pair;
  • the drug tumor suppressor sensitivity prediction unit predicts the IC50 value of the candidate drug in the cancer cell line environment by inputting the drug-cancer cell line instance pair into the drug tumor suppressor sensitivity prediction model.
  • the drug cancer inhibition sensitivity prediction device based on graph neural network is also provided with an input control interface at the input end of the processing module, which is used to control the input data; the output end of the processing module is also provided with an output control interface, which is used to control the output data; the output control interface is connected to the external interface, and its output expression form, in addition to the IC50 value, can also perform drug characteristic visualization output, cancer cell line characteristic visualization output, etc.
  • the relevant parts can refer to the partial description of the method embodiment.
  • the system embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application. A person of ordinary skill in the art can understand and implement it without creative work.
  • the present application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned drug cancer inhibition sensitivity prediction method based on graph neural network.
  • an electronic device comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned drug cancer inhibition sensitivity prediction method based on graph neural network.
  • a hardware structure diagram of any device with data processing capability in which the drug cancer inhibition sensitivity prediction method based on graph neural network provided in an embodiment of the present invention is located, in addition to the processor, memory and network interface shown in Figure 5, any device with data processing capability in which the device in the embodiment is located can also include other hardware according to the actual function of the device with data processing capability, which will not be described in detail.
  • the present application also provides a computer-readable storage medium on which computer instructions are stored, and when the instructions are executed by the processor, the above-mentioned drug cancer suppression sensitivity prediction method based on graph neural network is implemented.
  • the computer-readable storage medium can be an internal storage unit of any device with data processing capability described in any of the aforementioned embodiments, such as a hard disk or memory.
  • the computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), an SD card, a flash card (Flash Card), etc. equipped on the device.
  • the computer-readable storage medium can also include both an internal storage unit and an external storage device of any device with data processing capability.
  • the computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capability, and can also be used to temporarily store data that has been output or is to be output.

Landscapes

  • Chemical & Material Sciences (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computing Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

本发明公开了一种基于图神经网络的药物抑癌敏感性预测方法与装置，包括：获取候选药物及癌细胞系原始数据；提取候选药物的药物信息特征，并进行扩展；根据药物信息特征，构建药物分子特征向量，以药物的原子为节点，化学键为边，构建药物邻接矩阵、节点特征矩阵和边特征矩阵；构建并迭代图神经网络模型，得到药物图特征；药物图特征与药物分子特征向量聚合后，得到药物表征；获取存储基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量，提取癌细胞系表征；对药物表征和癌细胞系表征进行特征融合，得到药物-癌细胞系实例对，将实例对输入至药物抑癌敏感性预测模型中，预测得到候选药物在癌细胞系环境中的IC50值。

Description

一种基于图神经网络的药物抑癌敏感性预测方法与装置 技术领域
本发明主要涉及人工智能药物研发领域，具体涉及一种基于图神经网络的药物抑癌敏感性预测方法与装置。
背景技术
得益于深度学习算法的快速发展及大量癌细胞系及其全基因组信息的出现，以数据为驱动的癌症药物敏感性预测是加速癌症药物研发的必然技术趋势，能有效地缓解传统癌症药物研发周期长、成本高、效率低等问题，能为医生的临床用药提供关键指导，有助于癌症的个性化治疗。基于深度学习来进行药物抑癌敏感性预测是当下的热点，该类方法将药物和癌症细胞系数据输入到深度学习模型当中，组成“药物-癌细胞系”实例对，并通过深度学习算法来进行特征提取，挖掘实例对内在联系，从而进行药物抑癌敏感性预测。然而，当下的深度学习方法进行药物抑癌敏感性存在以下的局限性：
(1)药物数据获取麻烦、药物特征信息不全、药物特征稀疏，使得药物特征提取不够丰富。
(2)癌症细胞系表征困难，仅由少量癌细胞系组学数据表示会导致癌细胞系特征提取不够充分。
因此，亟需提出一种药物抑癌敏感性预测方法，来获取更丰富的药物和癌细胞系信息，提取更准确的“药物—癌细胞系”表征。
发明内容
针对现有技术的不足，本发明提出了一种基于图神经网络的药物抑癌敏感性预测方法与装置。
根据本发明实施例的第一方面，提供了一种基于图神经网络的药物抑癌敏感性预测方法，所述方法包括：
获取候选药物及癌细胞系原始数据；
提取候选药物的药物信息特征，并进行扩展；根据药物信息特征，构建药物分子特征向量，以药物的原子为节点，化学键为边，构建药物邻接矩阵、节点特征矩阵和边特征矩阵；
基于药物邻接矩阵、节点特征矩阵和边特征矩阵构建图神经网络模型，迭代图神经网络模型得到药物图特征；药物图特征与药物分子特征向量聚合后，得到药物表征；
获取基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量， 提取癌细胞系表征；
对药物表征和癌细胞系表征进行特征融合，得到药物-癌细胞系实例对；
将药物-癌细胞系实例对输入至药物抑癌敏感性预测模型中，预测得到候选药物在癌细胞系环境中的IC50值。
根据本发明实施例的第二方面，提供了一种基于图神经网络的药物抑癌敏感性预测装置，所述装置包括：多源数据整合模块、存储模块和处理模块；
多源数据整合模块，用于获取候选药物及癌细胞系原始数据；提取候选药物的药物信息特征，并进行扩展；根据药物信息特征，构建药物分子特征向量，以药物的原子为节点，化学键为边，构建药物邻接矩阵、节点特征矩阵和边特征矩阵；
存储模块包括：药物信息数据库和癌细胞系信息数据库；
药物信息数据库，用于存储药物分子特征向量、药物邻接矩阵、节点特征矩阵和边特征矩阵；
癌细胞系信息数据库，用于存储基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量；
处理模块包括：药物表征提取单元、癌细胞系表征提取单元、药物-癌细胞系实例对获取单元、药物抑癌敏感性预测单元；
药物表征提取单，基于药物邻接矩阵、节点特征矩阵和边特征矩阵构建图神经网络模型，迭代图神经网络模型得到药物图特征；药物图特征与药物分子特征向量聚合后，得到药物表征；
癌细胞系表征提取单元，通过获取基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量，提取癌细胞系表征；
药物-癌细胞系实例对获取单元，用于对药物表征和癌细胞系表征进行特征融合，得到药物-癌细胞系实例对；
药物抑癌敏感性预测单元，通过将药物-癌细胞系实例对输入至药物抑癌敏感性预测模型中，预测得到候选药物在癌细胞系环境中的IC50值。
根据本发明实施例的第三方面，提供了一种电子设备，包括一个或多个处理器，用于实现上述的基于图神经网络的药物抑癌敏感性预测方法。
根据本发明实施例的第四方面，提供了一种计算机可读存储介质，其上存储有程序，该程序被处理器执行时，用于实现上述的基于图神经网络的药物抑癌敏感性预测方法。
本发明的有益效果是：本发明提供了一种基于图神经网络的药物抑癌敏感性预测方法与装置，针对药物数据获取麻烦、药物特征信息不全、药物特征稀疏，提取候选药物的药物信 息特征，并进行扩展，融合了各大药物信息数据源，打破了单一数据源药物信息特征少、部分药物信息特征缺失等限制瓶颈，充分捕获药物特征信息，增加药物信息特征维度；针对癌细胞系表征困难，通过基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量来精准的表示癌细胞系，并通过Transformer网络和前馈神经网络对其进行编码，提取得到癌细胞系表征。本发明对药物表征和癌细胞系表征进行特征融合，得到药物-癌细胞系实例对，通过更准确、更丰富的药物-癌细胞系实例对，显著提高了药物抑癌敏感性的预测准确率。
附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的基于图神经网络的药物抑癌敏感性预测方法的流程图；
图2是本发明实施例提供的获取药物表征的流程性框图；
图3是本发明实施例提供的获取癌细胞系表征的流程性框图；
图4是本发明实施例提供的基于图神经网络的药物抑癌敏感性预测装置的示意图；
图5是本发明实施例提供的一种电子设备的示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
需要说明的是，在不冲突的情况下，下述的实施例及实施方式中的特征可以相互组合。
为了解决药物数据获取麻烦、药物特征信息不全、药物特征稀疏和癌细胞系表征困难的问题，实施例提供了一种基于图神经网络的药物抑癌敏感性预测方法与装置，融合了各大药物信息数据源，打破了单一数据源药物信息特征少、部分药物信息特征缺失等限制瓶颈，充分捕获药物特征信息，增加药物信息特征维度；通过基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量来精准的表示癌细胞系，并通过Transformer和前馈神经网络对其进行编码。本发明通过更准确、更丰富的“药物—癌细胞系”表征的提取，显著提高了药物抑癌敏感性的预测结果。
图1示出了本发明实施例提供的基于图神经网络的药物抑癌敏感性预测方法的流程图，所述方法具体包括以下步骤：
步骤S1，获取候选药物及癌细胞系原始数据。
步骤S2，提取候选药物的药物信息特征，并进行扩展；根据药物信息特征，构建药物分子特征向量，以药物的原子为节点，化学键为边，构建药物邻接矩阵、节点特征矩阵和边特征矩阵。
具体地，所述步骤S2具体包括以下子步骤：
步骤S201，以分子的SMILES序列为输入标准，从各药物信息数据源(DeepChem、RDKit、PubChem等)中获取药物信息特征，通过阅读各药物信息数据源官方文档，明确药物信息特征中每一维度的具体含义。
其中，可提取的药物信息特征包括药物的原子特征(原子类型、形式电荷、杂化轨道等)、药物的化学键特征(健类型、同环性、共轭性等)、药物分子特征(分子量、价电子数等)。
以化合物环丁烷为例，其SMILES序列为“C1CCC1”，通过“MolGraphConvFeaturizer”可以得到第一个“C”原子的原子特征为“[1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0]”，特征维度为30维，其中1-10维表示原子类型，11维表示形式电荷数，12-14维表示杂化轨道，15-16维表示氢键供体或受体，17维表示原子是否属于芳香环，18-24维表示原子的度，25-30维表示与原子相连的氢的数量。通过“MolGraphConvFeaturizer”可以得到第一个“C”原子和第二个“C”原子之间的连边的边特征为“[1,0,0,0,1,0,1,0,0,0,0]”,特征维度为11维，其中1-4维表示连边的健类型，5维表示边所连的两个原子是否在同一个环里，6维表示连边是否是共轭的，7-11维表示连边的键的立体构型。通过“rdkit”可以得到“C1CCC1”的分子特征如分子量为56.0626，忽略氢的分子平均分子量为48.0440，油水分配系数LogP为1.5604等。
步骤S202，撰写可提取的药物信息特征标准文档，对步骤S201中得到的药物信息特征进行去重、补全、融合。
步骤S203，根据步骤S202中得到的药物信息，以药物的原子为节点，化学键为边，构建药物邻接矩阵A、节点特征矩阵V、边特征矩阵E和药物分子特征向量Molecule_Feature。
由此，药物邻接矩阵A由维度为(shape，shape)的矩阵形式给出，其中“shape”代表该药物的节点数量(即药物的原子数量)。
节点特征矩阵V由维度为(shape，dim1)的矩阵形式给出，其中“dim1”代表该药物每个节点的特征数量(即药物每个原子的特征数量)。
边特征矩阵E由维度为(shape，shape，dim2)的邻接矩阵形式给出，其中“dim2”代表该药物每条边的特征数量(即药物化学键的特征数量)。进一步地，当两节点间不存在边时，边特征矩阵中相应的坐标元素用“0”填充。
药物分子特征向量Molecule_Feature由维度为(1，dim3)的特征向量形式给出，其中“的 特征3”表示该药物的分子特征维度。
对于药物邻接矩阵A，由于其是固定的，因此可以直接通过基准数据源得到。
对于节点特征矩阵V的构建与边特征矩阵E，由于两者的构建方法相同，以节点特征矩阵的构建为例进行阐述。
为表达方便，本实例只取两个数据源，并将数据源A和数据源B可以获得的节点特征矩阵VA，VB分别表示成以下形式：
若特征FB1并未出现在数据源A中，本实例将特征FB1直接添加到VA末尾，即可表示成以下形式：
若特征FB1与数据源A中的FA1是同一个性质的表征，当特征FA1与FB1是离散化数据时，本实例中取FA1与FB1中表示更多的那个特征作为该性质的表征。如原子的“Hybridization”属性，MolGraphConvFeaturizer提取器只将其表示成“sp”，“sp2”，“sp3”，而DMPNNFeaturizer提取器将其表示成“sp”，“sp2”，“sp3”，“sp3d”，“sp3d2”，因此对于“Hybridization”属性，本实例采用DMPNNFeaturizer提取器中提取到的属性作为最终属性；当特征FA1与FB1是非离散化数据时，本实例将FA1与FB1直接相连作为该性质的表征，即可表示成以下形式：
对于药物分子特征向量Molecule_Feature，可由不同数据源得到不同的分子特征，并通过“one-hot”等编码方式，将这些分子特征编码成药物分子特征向量Molecule_Feature。
步骤S204，将药物邻接矩阵A，节点特征矩阵V，边特征矩阵E以及药物分子特征向量Molecule_Feature传入到药物信息数据库中进行存储。
如图2所示，步骤S3，基于药物邻接矩阵、节点特征矩阵和边特征矩阵构建图神经网络模型，迭代图神经网络模型得到药物图特征；药物图特征与药物分子特征向量聚合后，得到药物表征。
需要说明的是，在本实例中可以从从药物信息数据库中获取物信息特征及相应的药物邻接矩阵A，节点特征矩阵V，边特征矩阵E等。搭建并迭代图神经网络模型，以节点更新、边更新的方式进行消息传递和聚合，得到药物图特征(Graph_Feature)。与药物信息数据库中的药物分子特征向量(Molecule_Feature)聚合后，得到最终的药物表征(Whole_Feature)。
具体地，所述步骤S3具体包括以下子步骤：
步骤S301，以搭建并迭代图神经网络模型，以节点更新、边更新的方式进行消息传递和聚合，得到药物图特征(Graph_Feature)。
其中，以节点更新的方式更新图神经网络模型，在图神经网络模型第l层进行节点特征更新，表达式如下：
式中，边特征矩阵为维度为(shape，shape，dim2)的矩阵，其中“dim2”代表药物每条边的特征数量，即药物化学键的特征数量；在节点更新过程中，将边特征矩阵拆分成“dim2”个(shape，shape，1)的边特征矩阵；V(l+1)表示第l+1层的节点特征矩阵,Updata表示更新函数,V(l)表示第l层的节点特征子矩阵，σ为激活函数，表示第l层上的边特征矩阵经过拆分后的第i个边特征子矩阵，Wi (l+1)和表示图神经网络模型参数。
以边更新的方式更新图神经网络模型，在图神经网络模型第l层进行边特征更新，表达式如下：

式中，表示第l+1层的节点i与节点j之间的连边eij的更新变量，σ为激活函数，Vi (l+1)表示节点i在第l+1层的特征向量，Vj (l+1)表示节点j在第l+1层的特征向量，和表示图神经网络模型参数，表示第l+1层的节点i与节点j之间的边特征矩阵，Updata表示更新函数，表示第l层的节点i与节点j之间的边特征矩阵。
步骤S303，计算药物图特征(Graph_Feature)，表达式如下：
其中，|V|表示节点数量，Vi (L)表示节点i在第L层的特征向量，L为图神经网络模型的层数。
步骤S304，药物图特征与药物分子特征向量进行聚合，得到药物表征，表达式如下：
Whole_Feature＝Concat(Graph_Feature,Molecule_Feature)
式中，Whole_Feature表示药物表征，Graph_Feature表示药物图特征，Molecule_Feature表示药物分子特征。
如图3所示，步骤S4，获取基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量，提取癌细胞系表征。
具体地，所述步骤S4具体包括以下子步骤：
步骤S401，获取基因序列在癌细胞系中的各表达量的相关信息，对于癌细胞系C1，其表达序列Xc1可以表达成以下形式：

其中，n表示为癌细胞系中致病基因数量，Xc1,gi表示第i个致病基因在癌细胞系中基因表达(expressc1,gi)、基因突变(mutatc1,gi)、基因甲基化(methyc1,gi)、基因复制水平copyc1,gi的表达量。
步骤S402，将癌细胞系C1的表达序列Xc1输入至Transformer网络中，输出表征向量Zc1，表达式如下：

特别地，本实例在Transformer网络中采用多头注意力机制。其中，表示癌细胞系C1的表达序列Xc1在经过第i个注意力头后得到的表征，Softmax为激活函数，分别为Transformer网络的Q，K，V矩阵，Zc1为Xc1经过Transformer网络后的输出表征。
步骤S403，将表征向量Zc1输入至前馈神经网络中，得到癌细胞系表征Cell_Feature，表达式如下：
Cell_Feature＝σ(Zc1W+b)
式中，σ为激活函数，W和b表示前馈神经网络模型参数
步骤S5，对药物表征和癌细胞系表征进行特征融合，得到药物-癌细胞系实例对，表达式如下：
Mol_Cell_Feature＝Whole_Feature||Cell_Feature
式中，Mol_Cell_Feature表示药物-癌细胞系实例对，Whole_Feature表示药物表征，Cell_Feature表示癌细胞系表征。
步骤S6，将药物-癌细胞系实例对输入至药物抑癌敏感性预测模型中，预测得到候选药物在癌细胞系环境中的IC50值。
进一步地，药物抑癌敏感性预测模型为一分类器；将将药物-癌细胞系实例对输入至分类器中，进行药物抑癌敏感性预测，预测得到药物在癌细胞系环境中的IC50值，由此量化药物对癌细胞系的抑制能力大小，表达式如下：
Pred＝Softmax(Mol_Cell_Feature*W+b)
式中，Mol_Cell_Feature表示药物-癌细胞系实例对，W和b为药物抑癌敏感性预测模型参数。
综上所述，本发明提供了一种基于图神经网络的药物抑癌敏感性预测方法，针对药物数据获取麻烦、药物特征信息不全、药物特征稀疏，提取候选药物的药物信息特征，并进行扩展，融合了各大药物信息数据源，打破了单一数据源药物信息特征少、部分药物信息特征缺失等限制瓶颈，充分捕获药物特征信息，增加药物信息特征维度；针对癌细胞系表征困难，通过基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量来精准的表示癌细胞系，并通过Transformer网络和前馈神经网络对其进行编码，提取得到癌细胞系表征。本发明对药物表征和癌细胞系表征进行特征融合，得到药物-癌细胞系实例对，通过更准确、更丰富的药物-癌细胞系实例对，显著提高了药物抑癌敏感性的预测准确率。
本发明实施例还提供了一种基于图神经网络的药物抑癌敏感性预测装置，如图4所示，所述装置包括：多源数据整合模块、存储模块和处理模块；
多源数据整合模块，用于获取候选药物及癌细胞系原始数据；提取候选药物的药物信息特征，并进行扩展；根据药物信息特征，构建药物分子特征向量，以药物的原子为节点，化学键为边，构建药物邻接矩阵、节点特征矩阵和边特征矩阵；
存储模块包括：药物信息数据库和癌细胞系信息数据库；
药物信息数据库，用于存储药物分子特征向量、药物邻接矩阵、节点特征矩阵和边特征矩阵；
癌细胞系信息数据库，用于存储基因序列在癌细胞系中基因表达、基因突变、基因甲基 化、基因拷贝数的表达量；
处理模块包括：药物表征提取单元、癌细胞系表征提取单元、药物-癌细胞系实例对获取单元、药物抑癌敏感性预测单元；
药物表征提取单，基于药物邻接矩阵、节点特征矩阵和边特征矩阵构建图神经网络模型，迭代图神经网络模型得到药物图特征；药物图特征与药物分子特征向量聚合后，得到药物表征；
癌细胞系表征提取单元，通过获取基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量，提取癌细胞系表征；
药物-癌细胞系实例对获取单元，用于对药物表征和癌细胞系表征进行特征融合，得到药物-癌细胞系实例对；
药物抑癌敏感性预测单元，通过将药物-癌细胞系实例对输入至药物抑癌敏感性预测模型中，预测得到候选药物在癌细胞系环境中的IC50值。
需要说明的是，所述基于图神经网络的药物抑癌敏感性预测装置在处理模块的输入端还设置有输入控制接口，用于控制输入数据；处理模块的输出端还设置有输出控制接口，用于控制输出数据；输出控制接口与外部接口相连，其输出表现形式除IC50值外，还可进行药物特征可视化输、癌细胞系特征可视化输出等。
关于上述实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
对于装置实施例而言，由于其基本对应于方法实施例，所以相关之处参见方法实施例的部分说明即可。以上所描述的系统实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本申请方案的目的。本领域普通技术人员在不付出创造性劳动的情况下，即可以理解并实施。
相应的，本申请还提供一种电子设备，包括：一个或多个处理器；存储器，用于存储一个或多个程序；当所述一个或多个程序被所述一个或多个处理器执行，使得所述一个或多个处理器实现如上述的基于图神经网络的药物抑癌敏感性预测方法。如图5所示，为本发明实施例提供的基于图神经网络的药物抑癌敏感性预测方法所在任意具备数据处理能力的设备的一种硬件结构图，除了图5所示的处理器、内存以及网络接口之外，实施例中装置所在的任意具备数据处理能力的设备通常根据该任意具备数据处理能力的设备的实际功能，还可以包括其他硬件，对此不再赘述。
相应的，本申请还提供一种计算机可读存储介质，其上存储有计算机指令，该指令被处理器执行时实现如上述的基于图神经网络的药物抑癌敏感性预测方法。所述计算机可读存储介质可以是前述任一实施例所述的任意具备数据处理能力的设备的内部存储单元，例如硬盘或内存。所述计算机可读存储介质也可以是外部存储设备，例如所述设备上配备的插接式硬盘、智能存储卡(Smart Media Card，SMC)、SD卡、闪存卡(Flash Card)等。进一步的，所述计算机可读存储介还可以既包括任意具备数据处理能力的设备的内部存储单元也包括外部存储设备。所述计算机可读存储介质用于存储所述计算机程序以及所述任意具备数据处理能力的设备所需的其他程序和数据，还可以用于暂时地存储已经输出或者将要输出的数据。
以上实施例仅用于说明本发明的设计思想和特点，其目的在于使本领域内的技术人员能够了解本发明的内容并据以实施，本发明的保护范围不限于上述实施例。所以，凡依据本发明所揭示的原理、设计思路所作的等同变化或修饰，均在本发明的保护范围之内。

Claims (10)

  1. 一种基于图神经网络的药物抑癌敏感性预测方法，其特征在于，所述方法包括：
    获取候选药物及癌细胞系原始数据；
    提取候选药物的药物信息特征，并进行扩展；根据药物信息特征，构建药物分子特征向量，以药物的原子为节点，化学键为边，构建药物邻接矩阵、节点特征矩阵和边特征矩阵；
    基于药物邻接矩阵、节点特征矩阵和边特征矩阵构建图神经网络模型，迭代图神经网络模型得到药物图特征；药物图特征与药物分子特征向量聚合后，得到药物表征；
    获取基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量，提取癌细胞系表征；
    对药物表征和癌细胞系表征进行特征融合，得到药物-癌细胞系实例对；
    将药物-癌细胞系实例对输入至药物抑癌敏感性预测模型中，预测得到候选药物在癌细胞系环境中的IC50值。
  2. 根据权利要求1所述的基于图神经网络的药物抑癌敏感性预测方法，其特征在于，药物信息特征包括药物的原子特征、药物的化学键特征、药物分子特征；
    药物邻接矩阵为维度为(shape，shape)的矩阵，其中“shape”代表药物的节点数量，即药物的原子数量；
    节点特征矩阵为维度为(shape，dim1)的矩阵，其中“dim1”代表药物每个节点的特征数量，即药物每个原子的特征数量；
    边特征矩阵为维度为(shape，shape，dim2)的矩阵，其中“dim2”代表药物每条边的特征数量，即药物化学键的特征数量；
    药物分子特征向量为维度为(1，dim3)的特征向量，其中“dim3”代表药物的分子特征维度。
  3. 根据权利要求1或2所述的基于图神经网络的药物抑癌敏感性预测方法，其特征在于，节点特征矩阵与边特征矩阵的构建方法相同；
    其中，节点特征矩阵的构建过程包括：
    获取数据源A和数据源B，数据源A和数据源B对应的第一节点特征矩阵VA、第二节点特征矩阵VB的表达式如下：
    当特征FB1未出现在数据源A中，将特征FB1直接添加到第一节点特征矩阵VA末尾，得到节点特征矩阵V，表达式如下：
    当特征FB1与数据源A中的FA1是同一个性质的表征，且特征FA1与FB1是离散化数据时，取FA1与FB1中表示更多的特征作为该性质的表征；
    当特征FA1与FB1是非离散化数据时，将FA1与FB1直接相连作为该性质的表征，节点特征矩阵V的表达式为：
  4. 根据权利要求1所述的基于图神经网络的药物抑癌敏感性预测方法，其特征在于，迭代图神经网络模型得到药物图特征包括：
    以节点更新的方式更新图神经网络模型，其中在图神经网络模型第l层进行节点特征更新，表达式如下：
    式中，边特征矩阵为维度为(shape，shape，dim2)的矩阵，其中“dim2”代表药物每条边的特征数量，即药物化学键的特征数量；在节点更新过程中，将边特征矩阵拆分成“dim2”个(shape，shape，1)的边特征矩阵；V(l+1)表示第l+1层的节点特征矩阵,Updata表示更新函数,V(l)表示第l层的节点特征子矩阵，σ为激活函数，表示第l层上的边特征矩阵经过拆分后的第i个边特征子矩阵，Wi (l+1)和表示图神经网络模型参数。
  5. 根据权利要求1所述的基于图神经网络的药物抑癌敏感性预测方法，其特征在于，迭 代图神经网络模型得到药物图特征包括：
    以边更新的方式更新图神经网络模型，其中在图神经网络模型第l层进行边特征更新，表达式如下：

    式中，表示第l+1层的节点i与节点j之间的连边eij的更新变量，σ为激活函数，Vi (l+1)表示节点i在第l+1层的特征向量，Vj (l+1)表示节点j在第l+1层的特征向量，和表示图神经网络模型参数，表示第l+1层的节点i与节点j之间的边特征矩阵，Updata表示更新函数，表示第l层的节点i与节点j之间的边特征矩阵。
  6. 根据权利要求1所述的基于图神经网络的药物抑癌敏感性预测方法，其特征在于，获取存储基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量，提取癌细胞系表征包括：
    癌细胞系C1的表达序列Xc1的表达式如下：

    式中，n表示癌细胞系中致病基因数量，Xc1,gi表示第i个致病基因在癌细胞系中的表达量，expressc1,gi表示第i个致病基因在癌细胞系中的基因表达的表达量，mutatc1,gi表示第i个致病基因在癌细胞系中的基因突变的表达量，methyc1,gi表示第i个致病基因在癌细胞系中的基因甲基化的表达量，copyc1,gi表示第i个致病基因在癌细胞系中的基因拷贝数的表达量；
    将癌细胞系C1的表达序列Xc1输入至Transformer网络中，输出表征向量Zc1；
    将表征向量Zc1输入至前馈神经网络中，得到癌细胞系表征。
  7. 根据权利要求1所述的基于图神经网络的药物抑癌敏感性预测方法，其特征在于，将药物-癌细胞系实例对输入至药物抑癌敏感性预测模型中，预测得到候选药物在癌细胞系环境中的IC50值包括：
    药物抑癌敏感性预测模型为一分类器；将将药物-癌细胞系实例对输入至分类器中，进行药物抑癌敏感性预测，预测得到药物在癌细胞系环境中的IC50值，表达式如下：
    Pred＝Softmax(Mol_Cell_Feature*W+b)
    式中，Mol_Cell_Feature表示药物-癌细胞系实例对，W和b为药物抑癌敏感性预测模型参 数。
  8. 一种基于图神经网络的药物抑癌敏感性预测装置，其特征在于，所述装置包括：多源数据整合模块、存储模块和处理模块；
    多源数据整合模块，用于获取候选药物及癌细胞系原始数据；提取候选药物的药物信息特征，并进行扩展；根据药物信息特征，构建药物分子特征向量，以药物的原子为节点，化学键为边，构建药物邻接矩阵、节点特征矩阵和边特征矩阵；
    存储模块包括：药物信息数据库和癌细胞系信息数据库；
    药物信息数据库，用于存储药物分子特征向量、药物邻接矩阵、节点特征矩阵和边特征矩阵；
    癌细胞系信息数据库，用于存储基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量；
    处理模块包括：药物表征提取单元、癌细胞系表征提取单元、药物-癌细胞系实例对获取单元、药物抑癌敏感性预测单元；
    药物表征提取单，基于药物邻接矩阵、节点特征矩阵和边特征矩阵构建图神经网络模型，迭代图神经网络模型得到药物图特征；药物图特征与药物分子特征向量聚合后，得到药物表征；
    癌细胞系表征提取单元，通过获取基因序列在癌细胞系中基因表达、基因突变、基因甲基化、基因拷贝数的表达量，提取癌细胞系表征；
    药物-癌细胞系实例对获取单元，用于对药物表征和癌细胞系表征进行特征融合，得到药物-癌细胞系实例对；
    药物抑癌敏感性预测单元，通过将药物-癌细胞系实例对输入至药物抑癌敏感性预测模型中，预测得到候选药物在癌细胞系环境中的IC50值。
  9. 一种电子设备，其特征在于，包括一个或多个处理器，用于实现权利要求1-7中任一项所述的基于图神经网络的药物抑癌敏感性预测方法。
  10. 一种计算机可读存储介质，其上存储有程序，其特征在于，该程序被处理器执行时，用于实现权利要求1-7中任一项所述的基于图神经网络的药物抑癌敏感性预测方法。
PCT/CN2023/098718 2023-06-06 2023-06-06 一种基于图神经网络的药物抑癌敏感性预测方法与装置 Ceased WO2024250183A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2023/098718 WO2024250183A1 (zh) 2023-06-06 2023-06-06 一种基于图神经网络的药物抑癌敏感性预测方法与装置

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2023/098718 WO2024250183A1 (zh) 2023-06-06 2023-06-06 一种基于图神经网络的药物抑癌敏感性预测方法与装置

Publications (1)

Publication Number Publication Date
WO2024250183A1 true WO2024250183A1 (zh) 2024-12-12

Family

ID=93794721

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/098718 Ceased WO2024250183A1 (zh) 2023-06-06 2023-06-06 一种基于图神经网络的药物抑癌敏感性预测方法与装置

Country Status (1)

Country Link
WO (1) WO2024250183A1 (zh)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160246919A1 (en) * 2013-10-08 2016-08-25 The Regents Of The University Of California Predictive optimization of network system response
CN108830040A (zh) * 2018-06-07 2018-11-16 中南大学 一种基于细胞系和药物相似性网络的药物敏感性预测方法
CN113782089A (zh) * 2021-11-15 2021-12-10 浙江大学 基于多组学数据融合的药物敏感性预测方法和装置
CN114496303A (zh) * 2022-01-06 2022-05-13 湖南大学 一种基于多通道神经网络的抗癌药物筛选方法
CN114974610A (zh) * 2022-04-14 2022-08-30 湖南大学 基于图神经网络和自动编码器的抗癌药物敏感性预测方法
WO2022214036A1 (zh) * 2021-04-09 2022-10-13 至本医疗科技（上海）有限公司 用于预测药物敏感状态的方法、设备和存储介质
CN115274136A (zh) * 2022-08-26 2022-11-01 上海交通大学 整合多组学与必需基因的肿瘤细胞系药物响应预测方法
CN116110509A (zh) * 2022-11-15 2023-05-12 浙江大学 基于组学一致性预训练的药物敏感性预测方法和装置

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160246919A1 (en) * 2013-10-08 2016-08-25 The Regents Of The University Of California Predictive optimization of network system response
CN108830040A (zh) * 2018-06-07 2018-11-16 中南大学 一种基于细胞系和药物相似性网络的药物敏感性预测方法
WO2022214036A1 (zh) * 2021-04-09 2022-10-13 至本医疗科技（上海）有限公司 用于预测药物敏感状态的方法、设备和存储介质
CN113782089A (zh) * 2021-11-15 2021-12-10 浙江大学 基于多组学数据融合的药物敏感性预测方法和装置
CN114496303A (zh) * 2022-01-06 2022-05-13 湖南大学 一种基于多通道神经网络的抗癌药物筛选方法
CN114974610A (zh) * 2022-04-14 2022-08-30 湖南大学 基于图神经网络和自动编码器的抗癌药物敏感性预测方法
CN115274136A (zh) * 2022-08-26 2022-11-01 上海交通大学 整合多组学与必需基因的肿瘤细胞系药物响应预测方法
CN116110509A (zh) * 2022-11-15 2023-05-12 浙江大学 基于组学一致性预训练的药物敏感性预测方法和装置

Similar Documents

Publication Publication Date Title
CN113782089B (zh) 基于多组学数据融合的药物敏感性预测方法和装置
CN112990385B (zh) 一种基于半监督变分自编码器的主动众包图像学习方法
CN111476038A (zh) 长文本生成方法、装置、计算机设备和存储介质
WO2025065954A1 (zh) 植物表型预测
CN111325340B (zh) 信息网络关系预测方法及系统
CN115146168A (zh) 一种融合知识图谱和图对比学习的图书推荐方法及系统
CN110968692B (zh) 一种文本分类方法及系统
CN111291618A (zh) 标注方法、装置、服务器和存储介质
CN117056597A (zh) 基于噪声增强的对比学习图推荐方法
CN112508181A (zh) 一种基于多通道机制的图池化方法
Chen et al. AI-driven accelerated discovery of intercalation-type cathode materials for magnesium batteries
Wu et al. KGETCDA: an efficient representation learning framework based on knowledge graph encoder from transformer for predicting circRNA-disease associations
Liu et al. Semantic relation-aware graph attention network with noise augmented layer-wise contrastive learning for recommendation
CN116705194B (zh) 一种基于图神经网络的药物抑癌敏感性预测方法与装置
CN108614932B (zh) 基于边图的线性流重叠社区发现方法、系统及存储介质
Zheng et al. A front-end fusion feature-based machine learning engine for rechargeable battery manufacturing accelerates screening of organic electrodes
CN116204628A (zh) 一种知识图谱增强的物流知识神经协同过滤推荐方法
CN121306588A (zh) 一种深度张量分解与超图卷积用于预测miRNA-lncRNA-疾病三元关联方法
Wang et al. POMP: Pathology-omics Multimodal Pre-training Framework for Cancer Survival Prediction
CN116227592A (zh) 一种多源知识图谱对齐模型及其构建方法、装置及应用
CN116804997A (zh) 基于图神经网络的中文相似案例推荐方法、装置和设备
Kim et al. Machine Learning-Assisted Crystal Structure Prediction of Solid-State Electrolytes Reveals Superior Ionic Conductivity in Metastable Edge-Sharing Phases
CN106407403A (zh) 基于学习的受约束服务聚合排名预测方法及装置
Lai et al. Estimation and variable selection for generalised partially linear single-index models
CN112749297B (zh) 视频推荐方法、装置、计算机设备和计算机可读存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23940073

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE