CN109887540A - A kind of drug targets interaction prediction method based on heterogeneous network insertion - Google Patents
A kind of drug targets interaction prediction method based on heterogeneous network insertion Download PDFInfo
- Publication number
- CN109887540A CN109887540A CN201910035394.7A CN201910035394A CN109887540A CN 109887540 A CN109887540 A CN 109887540A CN 201910035394 A CN201910035394 A CN 201910035394A CN 109887540 A CN109887540 A CN 109887540A
- Authority
- CN
- China
- Prior art keywords
- drug
- target
- heterogeneous network
- vector
- interaction
- 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.)
- Pending
Links
- 239000003596 drug target Substances 0.000 title claims abstract description 92
- 230000003993 interaction Effects 0.000 title claims abstract description 66
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000003780 insertion Methods 0.000 title claims abstract description 16
- 230000037431 insertion Effects 0.000 title claims abstract description 16
- 239000003814 drug Substances 0.000 claims abstract description 115
- 229940079593 drug Drugs 0.000 claims abstract description 106
- 239000013598 vector Substances 0.000 claims abstract description 58
- 238000013508 migration Methods 0.000 claims abstract description 22
- 230000005012 migration Effects 0.000 claims abstract description 22
- 238000007637 random forest analysis Methods 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims description 52
- 239000011159 matrix material Substances 0.000 claims description 28
- 238000003062 neural network model Methods 0.000 claims description 8
- 238000005295 random walk Methods 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 6
- 230000001052 transient effect Effects 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 8
- 238000013145 classification model Methods 0.000 abstract description 3
- 230000002452 interceptive effect Effects 0.000 abstract description 3
- 239000000126 substance Substances 0.000 abstract description 3
- 238000012795 verification Methods 0.000 abstract description 3
- 238000013528 artificial neural network Methods 0.000 abstract description 2
- 238000012360 testing method Methods 0.000 description 8
- 238000010801 machine learning Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 4
- 238000012827 research and development Methods 0.000 description 4
- 102000003688 G-Protein-Coupled Receptors Human genes 0.000 description 3
- 108090000045 G-Protein-Coupled Receptors Proteins 0.000 description 3
- 102000004310 Ion Channels Human genes 0.000 description 3
- 238000003012 network analysis Methods 0.000 description 3
- 239000002547 new drug Substances 0.000 description 3
- 102000004190 Enzymes Human genes 0.000 description 2
- 108090000790 Enzymes Proteins 0.000 description 2
- 108020005497 Nuclear hormone receptor Proteins 0.000 description 2
- 238000004617 QSAR study Methods 0.000 description 2
- 238000009412 basement excavation Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 102000006255 nuclear receptors Human genes 0.000 description 2
- 108020004017 nuclear receptors Proteins 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011524 similarity measure Methods 0.000 description 2
- 108090000862 Ion Channels Proteins 0.000 description 1
- 125000003275 alpha amino acid group Chemical group 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 229940000406 drug candidate Drugs 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000003777 experimental drug Substances 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012966 insertion method Methods 0.000 description 1
- 239000003446 ligand Substances 0.000 description 1
- 238000003032 molecular docking Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 102000004169 proteins and genes Human genes 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of drug targets interaction prediction methods based on heterogeneous network insertion, this method is tended to the interaction of similar target based on the similar drug of chemical structure it is assumed that drug-drug similitude network, target-target similitude network and drug-target interactive network are merged into drug-target heterogeneous network；It using the migration sequence based on start node, constructs neural network classification model and is inputted migration sequence as it, disaggregated model is trained and the vector for learning to obtain all nodes indicates；For drug-target interaction prediction, given a pair of drug-target pair, the vector that corresponding drug and target are extracted from the knot vector that study obtains indicates, and Hadamard product operation is carried out to two vectors, using obtained result as the input of random forest grader, final prediction result is obtained.According to experimental verification it is found that this method prediction effect and applicability are preferable.
Description
Technical field
The invention belongs to drug prediction and analysis fields, and in particular to a kind of drug targets phase based on heterogeneous network insertion
Interaction prediction technique.
Background technique
Medicament research and development technology is rapidly developed in nearly 30 years, including genomics, proteomics and is
Multiple means including system biology are widely used to the identification of drug and target and the research and development of original new drug.But, it grinds
The drug of hair brand new still remains the problem that costly, risk is higher, the period is very long and success rate is very low.Currently,
The drug that a brand new is researched and developed in success averagely needs the time of 10-15 and the investment of multi-million dollar.Due to completely newly tying
The drug of structure often has the side effect for being difficult to predict, and about 90% experimental drug can not pass through a clinical trial phase.It faces
The Innovation Input of original novel drugs increases year by year in the world and risk of failure increases year by year, and new drug quantity approved stagnates,
The cost, risk and period of medicament research and development can be greatly reduced, and the drug reorientation of known drug new application can be excavated already
The strategy to become more and more important as medicament research and development.However how to be found to have from a large amount of unknown drug-target relationships pair potential
Interaction be drug reorientation problem research focus.
Traditional drug-target interaction prediction method is roughly divided into based on ligand and based on two methods of structure, example
Such as modeling of quantitative structure activity relationship (QSAR), pharmacophore, molecular docking.In recent years, occur various using large-scale
Common data base such as DrugBank, PubChem and ChEMBL etc. have come the method for predicting the interaction of drug-target based on net
Network analysis, also have based on machine learning.It, can be by medicine in the mutual prediction technique of drug-target of Excavation Cluster Based on Network Analysis
Object and target interacting space are considered as a bipartite graph, and node therein is drug and target, and side therein indicates corresponding
Drug-target Thermodynamic parameters are it is known that the new drug of prediction-target interaction finds missing to being equivalent in bipartite graph
Side；Drug-target interaction matrix can also be considered as incidence matrix, be predicted by the method for such as matrix decomposition potential
Interaction.In the drug based on machine learning-target interaction prediction method, the feature vector of drug and target is made
For input, interaction is used as label, the class label that potential drug-target interaction can be come out by model prediction
It determines.Excavation Cluster Based on Network Analysis and drug-target interaction prediction method based on machine learning have respective advantage and disadvantage, base
Known reachable path is too dependent in the computation model of network, and the model prediction accuracy very great Cheng based on machine learning
Whether the feature that the drug and target extracted is depended on degree is accurate and selection of negative sample also affects to a certain extent
Prediction accuracy.
Summary of the invention
(1) technical problems to be solved
Based on this, the invention proposes a kind of methods that network and machine learning combine, and specifically one kind is based on
The method of the drug and target Interaction Predicting of heterogeneous network insertion, this method can be mentioned from drug-target heterogeneous network
Take and learn to obtain the feature vector of drug and target, and be effectively predicted with machine learning method, and its prediction effect and
Applicability is good.
(2) technical solution
In order to solve the above technical problems, the invention proposes a kind of drug targets interactions based on heterogeneous network insertion
Prediction technique, comprising the following steps:
Step 1: obtaining the related data of drug and target as initial data；
Step 2: building drug-target heterogeneous network, respectively drug similarity matrix and the setting of target similarity matrix
Threshold parameter, for each pair of drug-drug to and target-target pair, when the drug-drug pair or target-target pair
Similarity be greater than specified threshold when, then add a line into drug-target heterogeneous network, the information on side include source node,
Destination node, source node type, destination node type, connection type and weight, wherein the weight on side is similarity；For medicine
Object target interaction matrix, only comprising 0 and 1 two value in matrix, 1 indicate corresponding drug and target there are it is known mutually
Effect, 0 indicates to interact unknown, and the known drug-target that there is interaction in part will be added to drug-to side is considered as
In target heterogeneous network, the information on side equally also includes source node, destination node, source node type, destination node type, connection
Type and weight, wherein the weight on side is 1, is finally obtained comprising drug to drug link information, target to target link information
With drug-target heterogeneous network of some drugs to target link information；
Step 3: the drug-target heterogeneous network based on Heterogeneous Information generates multiple migration sequences using random walk
Column generate training data according to migration sequence, training data are put into training in neural network model, after training is completed, i.e.,
The vector that all nodes in network can be obtained indicates；
Step 4: building drug-target pair sample set will be corresponded to for each pair of drug-target to as a sample
Drug vector sum target vector be Hadamard product operation obtain input of the new vector as sample, corresponding drug and target
Mark is then the label of sample with the presence or absence of interaction, if drug and target have interaction, label value is 1, otherwise
Value is 0；
Step 5: establishing final drug-target Interaction Predicting model using random forest disaggregated model.
Preferably, in the step 1, the drug and the related data of target are specifically that drug similitude, target are similar
Property and drug-target interact data.
Preferably, in the step 2, link information is extracted from drug similarity matrix and target similarity matrix respectively
When, set drug-drug to and target-target pair threshold parameter be respectively 0.2 and 0.3.
Preferably, in the step 2, in the building of the drug based on link information-target heterogeneous network, the side in network
It is two-way side, and each edge has weight, similitude of the weight on drug-drug side between drug, target-target side
Similitude of the weight between target, drug-target while and weight when target-drug be 1, each edge includes source section
The information such as point, destination node, source node type, destination node type, connection type and weight.
Preferably, in the step 3, the drug-target heterogeneous network based on Heterogeneous Information is raw using random walk
At multiple migration sequences, training data is generated according to migration sequence, training data is put into training in neural network model, training
Resulting network weight after completion is that the vector of all nodes in network indicates；All knot vector dimensions learnt
For n dimension, wherein n is positive integer.
Preferably, in the step 4, the sample input of the sample set is that corresponding drug vector sum target vector is done
Hadamard product operation obtains new vector, and it is constant that obtained vector dimension remains n dimension, wherein being per one-dimensional numerical value
Drug vector sum target vector corresponds to the product of the numerical value of dimension, and wherein n is positive integer.
Preferably, in the step 5, the training set that the drug-target Interaction Predicting model uses is from total
Training is concentrated with the stochastical sampling put back to and obtains, in the node of each tree of the training random forest disaggregated model, from institute
Have in feature according to a certain percentage randomly without a certain amount of feature of extraction put back to.
In addition, the invention also discloses a kind of electronic equipment, comprising:
At least one processor；And
At least one processor being connect with the processor communication, in which:
The memory is stored with the program instruction that can be executed by the processor, and the processor calls described program to refer to
It enables and is able to carry out the drug targets interaction prediction method described in any of the above embodiments based on heterogeneous network insertion.
In addition, the non-transient computer is readable the invention also discloses a kind of non-transient computer readable storage medium
Storage medium stores computer instruction, and it is described in any of the above embodiments based on isomery that the computer instruction executes the computer
The drug targets interaction prediction method of internet startup disk.
(3) beneficial effect
Compared with prior art, it is past based on the similar drug of chemical structure that the invention has the following beneficial effects: this method
Toward can with similar target interact it is assumed that by drug-drug similitude network, target-target similitude network and
Drug-target interactive network is merged into drug-target heterogeneous network；It is generated using random walk model with every in network
A node is the migration sequence of start node, constructs neural network classification model and inputs migration sequence as it, to classification
Model is trained and learns to obtain the vector expression of all nodes；For drug-target interaction prediction, a pair is given
Drug-target pair extracts the vector expression of corresponding drug and target from the knot vector that study obtains, and to two vectors
It carries out Hadamard product operation and obtains final prediction result using obtained result as the input of random forest grader.This
The prediction technique of invention can quickly be extracted from drug-target heterogeneous network and learn to obtain the feature of drug and target to
Amount, and predicted with machine learning method, and its prediction effect and applicability are good, are adapted for the data processing of batch.
Detailed description of the invention
The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, and attached drawing is schematically without that should manage
Solution is carries out any restrictions to the present invention, in the accompanying drawings:
Fig. 1 is the flow chart of prediction technique of the present invention；
Fig. 2 is the AUC value comparison diagram of each method on different data sets.
Specific embodiment
In order to make the foregoing objectives, features and advantages of the present invention clearer and more comprehensible, with reference to the accompanying drawing to the present invention
Specific embodiment be described in detail.Many details are explained in the following description in order to fully understand this hair
It is bright.But the invention can be embodied in many other ways as described herein, those skilled in the art can be not
Similar improvement is done in the case where violating intension of the present invention, therefore the present invention is not limited to the specific embodiments disclosed below.
Embodiment 1
As shown in Figure 1, the invention particularly discloses a kind of drug targets Interaction Predictings based on heterogeneous network insertion
Method, which comprises the following steps:
Step 1: obtaining the related data of drug and target as initial data；
Step 2: building drug-target heterogeneous network, respectively drug similarity matrix and the setting of target similarity matrix
Threshold parameter, for each pair of drug-drug to and target-target pair, when the drug-drug pair or target-target pair
Similarity be greater than specified threshold when, then add a line into drug-target heterogeneous network, the information on side include source node,
Destination node, source node type, destination node type, connection type and weight, wherein the weight on side is similarity；For medicine
Object target interaction matrix, only comprising 0 and 1 two value in matrix, 1 indicate corresponding drug and target there are it is known mutually
Effect, 0 indicates to interact unknown, and the known drug-target that there is interaction in part will be added to drug-to side is considered as
In target heterogeneous network, the information on side equally also includes source node, destination node, source node type, destination node type, connection
Type and weight, wherein the weight on side is 1, is finally obtained comprising drug to drug link information, target to target link information
With drug-target heterogeneous network of some drugs to target link information；
Step 3: the drug-target heterogeneous network based on Heterogeneous Information generates multiple migration sequences using random walk
Column generate training data according to migration sequence, training data are put into training in neural network model, after training is completed, i.e.,
The vector that all nodes in network can be obtained indicates；
Step 4: building drug-target pair sample set will be corresponded to for each pair of drug-target to as a sample
Drug vector sum target vector be Hadamard product operation obtain input of the new vector as sample, corresponding drug and target
Mark is then the label of sample with the presence or absence of interaction, if drug and target have interaction, label value is 1, otherwise
Value is 0；
Step 5: establishing final drug-target Interaction Predicting model using random forest disaggregated model.
In the step 1 of wherein one embodiment, the drug is specifically that drug is similar with the related data of target
Property, target similitude and drug-target interacts data.
In the step 2 of wherein one embodiment, mentioned respectively from drug similarity matrix and target similarity matrix
When taking link information, set drug-drug to and the threshold parameter of target-target pair similarity be respectively 0.2 He
0.3。
In the wherein step 2 of one embodiment, in the building of the drug based on link information-target heterogeneous network, net
In network when being two-way, and each edge has weight, similitude of the weight on drug-drug side between drug, target
Similitude of mark-target side weight between target, drug-target while and weight when target-drug be 1, each edge
It include the information such as source node, destination node, source node type, destination node type, connection type and weight.
In the wherein step 3 of one embodiment, the drug-target heterogeneous network based on Heterogeneous Information, using with
Machine migration generates multiple migration sequences, generates training data according to migration sequence, training data is put into neural network model
Training, resulting network weight after training is completed are that the vector of all nodes in network indicates；All nodes learnt
Vector dimension is 90 dimensions, wherein vector dimension can also be tieed up for n, and wherein n is positive integer.
In the step 4 of wherein one embodiment, the sample input of the sample set is corresponding drug vector sum
Target vector does Hadamard product operation and obtains new vector, and it is constant that obtained vector dimension remains 90 dimensions, wherein per one-dimensional
Numerical value be the numerical value that drug vector sum target vector corresponds to dimension product.
In the step 5 of wherein one embodiment, training that the drug-target Interaction Predicting model uses
Collection is to be concentrated with the stochastical sampling put back to from total training to obtain, in the section of each tree of the training random forest disaggregated model
When point, according to a certain percentage randomly without a certain amount of feature of extraction put back to from all features.
In addition, the meter being stored in non-transient computer readable storage medium can be used in the above-mentioned prediction technique of the application
Calculation machine instructs or the computer program that is executed by processor is realized.
Embodiment 2
To make the present invention it is more readily appreciated that the existing angle from data processing is of the invention based on heterogeneous network insertion to describe
Drug targets interaction prediction method, specifically include following four part, wherein first to Part III be prediction side
The main part of method, Part IV are experimental verification parts.
One, data prediction
The data prediction part contains step 1 and step 2 in embodiment 1, reads in drug similarity matrix file,
Drug similarity threshold is set, with D_simi,jIndicate drug DiWith drug DjSimilarity, work as D_simi,jWhen greater than threshold value,
Then add a drug DiTo drug DjSide and a drug DjTo drug DiSide into drug-target heterogeneous network, side
Information includes source node, destination node, source node type, destination node type, connection type and weight, and wherein the weight on side is
Similarity.
Similarly, target similarity matrix file is read in, target similarity threshold is set, is similarly operated.
Read in drug targets interaction matrix file, only comprising 0 and 1 two value in matrix, 1 indicate corresponding drug and
For target there are known interaction, 0 indicates that its interaction is unknown.For the known drug-target that there is interaction
It is right, then add a drug to target while and a target to drug while into drug-target heterogeneous network, the information on side
It equally also include source node, destination node, source node type, destination node type, connection type and weight, the wherein weight on side
It is 1.For the validity of test method, the present invention is only added to the known drug-target pair 90% that there is interaction, remains
Under 10% positive sample as test set.
It finally obtains and is connected to drug link information, target to target link information with some drugs to target comprising drug
The drug of information-target heterogeneous network.
Two, the study that the knot vector based on internet startup disk indicates
The part contains the steps 3 in embodiment 1, in order to obtain all drug nodes in drug-target heterogeneous network
Indicate that present invention uses the expression learning frameworks based on Heterogeneous Information network with the feature vector of target node, such as
HIN2Vec frame.Learning process is divided into following two part:
(1) training data is generated based on random walk and negative sampling
It gives migration sequence length L and using each node as the migration number K of start node, uses random walk strategy time
It goes through drug-target heterogeneous network and obtains the K migration sequence using each node as start node, the length of each migration sequence is equal
For L.Training data is generated from all migration sequences, in training data each sample be four-tuple (x, y, r, L (x,
Y, r)), x and y are two nodes, and r is expressed as relationship type, and L (x, y, r) indicates to whether there is relationship between node x and node y
R, 0 indicates to be not present, and 1 indicates exist.For a positive sample (x, y, r, 1), by random replacement node x or node y
Any one, and keeps relationship r constant, generates negative sample (x ', y, r, 0) or (x, y ', r, 0).
(2) expression of training neural network model learning network interior joint
Construct neural network model, the input layer of model be corresponding three one-hot of node x, y and relationship r encode to
Amount.Model is two disaggregated models, that is, gives two nodes x, y, predicts to whether there is relationship r, therefore, model between node
Output layer be one 0 to 1 value, indicating the node x, y of input, there are the probability of relationship r.The training objective of model is maximum
The prediction probability for changing positive sample, minimizes the prediction probability of negative sample.In parameter matrix W, that is, network of model after the completion of training
The vector of each node indicates.
Three, drug targets Interaction Predicting
The part contains the step 4 and step 5 in embodiment 1,
(1) building of sample set
Each sample is a pair of of drug-target pair, and the purpose of the model in the present invention is exactly drug-target in forecast sample
Mark is to the presence or absence of interaction.Sample set is divided into two subsets of training set and test set, and training set is of the invention for training
Sorter model, test set are then used for the correctness and validity of test model.The present invention is by known drug-target phase interaction
Positive sample of 90% used as training set, remaining 10% positive sample as test set.The positive and negative sample proportion of training set
It is 1 to 13, the positive and negative sample proportion of test set and the positive and negative sample proportion of initial data are consistent.
Meanwhile each sample includes sample input and two parts of sample label.Sample input is drug-target in sample
The Hadamard product operation result to corresponding drug vector and target vector is marked, concrete operation form is as follows:
If the corresponding vector of drug is expressed as in sample:
Di=(Di,1,Di,2,…,Di,90)
The corresponding vector of target is expressed as:
Tj=(Tj,1,Tj,2,…,Tj,90)
Then corresponding Hadamard product operation result are as follows:
Di⊙Tj=(Di,1*Tj,1,Di,2*Tj,2,…,Di,90*Tj,90)
I.e. the input of sample is the new vector that drug vector sum target vector carries out that Hadamard product operation obtains, and sample
Label indicate corresponding drug-target pair interaction whether it is known that if drug and target interaction it is known that if
Label value is 1, and otherwise value is 0.
(2) building of random forest disaggregated model
Random forest realizes that simple, precision is high, training speed is fast, is capable of handling the data of higher dimensional, and due to two
The introducing (sample is random and feature is random) of stochastic behaviour, so that random forest has certain noise resisting ability, it is not easy to fall into
Enter over-fitting.Therefore present invention uses random forests as last drug-target Interaction Predicting model.It is random gloomy
Woods (Random Forest, abbreviation RF) is a kind of integrated study classification method based on decision-tree model, and core concept is logical
Feature sampling is crossed to reduce trained variance, improves integrated generalization ability.Random forest is made of more trees, for every
Tree, the training set used are to be concentrated with the stochastical sampling put back to from total training to obtain, which means that having in total training set
A little samples may repeatedly appear in the training set of one tree, it is also possible to from the training set for not appearing in one tree.In training
When the node of each tree, uses and be characterized in from all features according to a certain percentage randomly without the extraction put back to.Output
Classification with majority voting (i.e. it is all tree output classifications modes) form determine.The present invention targetedly selects to use
Training set trains random forest disaggregated model, then carrys out the classifying quality of test model using test set.
Four, experimental verification
In order to verify the validity of prediction technique of the present invention, in 4 data of the truthful data of drug targets interaction
(G-Protein Coupled Receptors (GPCR), Ion Channels (IC), Nuclear Receptors on collection
(NR), Enzymes (E)) be tested, and with other 5 methods (BLM-NII, WNN-GIP, NetLabRLS, CMF,
BRDTI it) is compared.The data set of drug targets interaction is published in from Yamanishi et al.
Bioinformatics chemically with the public data of the consolidated forecast in genome space drug-target interactive network
Collection, each data set include drug-target interaction binary matrix, and each value in matrix indicates corresponding drug
Whether the interaction between target is known.The Similarity measures of drug use the chemical structure based on compound
SIMCOMP algorithm calculates, and target Similarity measures are the normalization Smith-Waterman score of target protein amino acid sequence.
For the accuracy and validity of evaluation and foreca result, the present invention is compared using AUC value index.AUC value is
A standard of a disaggregated model quality is measured, AUC value is area under the line of ROC curve, and AUC value is bigger, classifier classification
Effect is better.The experimental result of AUC value is specifically as shown in Figure 2.
Figure it is seen that prediction technique of the invention is in GPCR, Ion Channels, Enzymes this three data
AUC value on collection is superior to other methods, and because its data volume is too small on Nuclear Receptors data set, fail
Study is indicated to accurate vector from drug-target heterogeneous network, and AUC value is caused to fail better than other methods, but overall
On performance be an advantage over other methods.It can be seen that prediction technique proposed by the present invention has good prediction effect.
Unless otherwise defined, all technical and scientific terms used herein and belong to technical field of the invention
The normally understood meaning of technical staff is identical.Term as used herein in the specification of the present invention is intended merely to description tool
The purpose of the embodiment of body, it is not intended that in the limitation present invention.Each technical characteristic of embodiment described above can carry out arbitrarily
Combination, for simplicity of description, it is not all possible to each technical characteristic in above-described embodiment combination be all described, so
And as long as there is no contradiction in the combination of these technical features, it all should be considered as described in this specification.
Although the embodiments of the invention are described in conjunction with the attached drawings, but those skilled in the art can not depart from this hair
Various modifications and variations are made in the case where bright spirit and scope, such modifications and variations are each fallen within by appended claims
Within limited range.
Claims (9)
1. a kind of drug targets interaction prediction method based on heterogeneous network insertion, which comprises the following steps:
Step 1: obtaining the related data of drug and target as initial data；
Step 2: building drug-target heterogeneous network, respectively drug similarity matrix and target similarity matrix given threshold
Parameter, for each pair of drug-drug to and target-target pair, when the drug-drug pair or target-target pair is similar
Property value be greater than specified threshold when, then add a line into drug-target heterogeneous network, the information on side includes source node, target
Node, source node type, destination node type, connection type and weight, wherein the weight on side is similarity；For medicine target
Interaction matrix is marked, is only worth comprising 0 and 1 two in matrix, 1 indicates corresponding drug and target, and there are known phase interactions
With 0 indicates to interact unknown, and the known drug-target that there is interaction in part will be added to drug-target to side is considered as
It marks in heterogeneous network, the information on side equally also includes source node, destination node, source node type, destination node type, connection class
Type and weight, wherein the weight on side be 1, finally obtain comprising drug to drug link information, target to target link information and
Drug-target heterogeneous network of some drugs to target link information；
Step 3: the drug-target heterogeneous network based on Heterogeneous Information generates multiple migration sequences, root using random walk
Training data is generated according to migration sequence, training data is put into training in neural network model can be obtained after training is completed
The vector of all nodes indicates in network；
Step 4: building drug-target pair sample set, for each pair of drug-target to as a sample, by corresponding medicine
Object vector sum target vector does Hadamard product operation and obtains input of the new vector as sample, and corresponding drug and target are
The no label that there is interaction then and be sample, if drug and target have interaction, label value is 1, otherwise value
It is 0；
Step 5: establishing final drug-target Interaction Predicting model using random forest disaggregated model.
2. the drug targets interaction prediction method according to claim 1 based on heterogeneous network insertion, feature exist
In in the step 1, the related data of the drug and target is specifically drug similitude, target similitude and drug-target
Interact data.
3. the drug targets interaction prediction method according to claim 1 based on heterogeneous network insertion, feature exist
In, in the step 2, when extracting link information from drug similarity matrix and target similarity matrix respectively, set medicine
Object-drug to and target-target pair threshold parameter be respectively 0.2 and 0.3.
4. the drug targets interaction prediction method according to claim 1 based on heterogeneous network insertion, feature exist
In, in the step 2, in the building of the drug based on link information-target heterogeneous network, in network when being two-way,
And each edge has weight, similitude of the weight on drug-drug side between drug, target-target side weight is target
Similitude between mark, drug-target while and weight when target-drug be 1, each edge includes source node, target section
The information such as point, source node type, destination node type, connection type and weight.
5. the drug targets interaction prediction method according to claim 1 based on heterogeneous network insertion, feature exist
In in the step 3, the drug-target heterogeneous network based on Heterogeneous Information generates multiple migration sequences using random walk
Column generate training data according to migration sequence, and training data is put into training in neural network model, gained after training is completed
Network weight, be all nodes in network vector indicate；All knot vector dimensions learnt are n dimension, and wherein n is
Positive integer.
6. the drug targets interaction prediction method according to claim 1 based on heterogeneous network insertion, feature exist
In in the step 4, the sample input of the sample set is that corresponding drug vector sum target vector does Hadamard product operation
New vector is obtained, it is constant that obtained vector dimension remains n dimension, wherein being drug vector sum target per one-dimensional numerical value
Vector corresponds to the product of the numerical value of dimension, and wherein n is positive integer.
7. the drug targets interaction prediction method according to claim 1 based on heterogeneous network insertion, feature exist
In in the step 5, the training set that the drug-target Interaction Predicting model uses is to be concentrated with to put from total training
Return stochastical sampling obtain, in the node of each tree of the training random forest disaggregated model, from all features according to
Certain proportion is randomly without a certain amount of feature of extraction put back to.
8. a kind of electronic equipment characterized by comprising
At least one processor；And
At least one processor being connect with the processor communication, in which:
The memory is stored with the program instruction that can be executed by the processor, and the processor calls described program to instruct energy
Enough execute the drug targets interaction prediction method as described in any one of claim 1 to 7 based on heterogeneous network insertion.
9. a kind of non-transient computer readable storage medium, which is characterized in that the non-transient computer readable storage medium is deposited
Computer instruction is stored up, it is as described in any one of claim 1 to 7 based on isomery that the computer instruction executes the computer
The drug targets interaction prediction method of internet startup disk.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910035394.7A CN109887540A (en) | 2019-01-15 | 2019-01-15 | A kind of drug targets interaction prediction method based on heterogeneous network insertion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910035394.7A CN109887540A (en) | 2019-01-15 | 2019-01-15 | A kind of drug targets interaction prediction method based on heterogeneous network insertion |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109887540A true CN109887540A (en) | 2019-06-14 |
Family
ID=66925985
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910035394.7A Pending CN109887540A (en) | 2019-01-15 | 2019-01-15 | A kind of drug targets interaction prediction method based on heterogeneous network insertion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109887540A (en) |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110310703A (en) * | 2019-06-25 | 2019-10-08 | 中国人民解放军军事科学院军事医学研究院 | Prediction technique, device and the computer equipment of drug |
| CN110415763A (en) * | 2019-08-06 | 2019-11-05 | 腾讯科技（深圳）有限公司 | The interaction prediction method of drug and target, device, equipment and storage medium |
| CN110534153A (en) * | 2019-08-30 | 2019-12-03 | 广州费米子科技有限责任公司 | Target prediction system and method based on deep learning |
| CN110598130A (en) * | 2019-09-30 | 2019-12-20 | 重庆邮电大学 | Movie recommendation method integrating heterogeneous information network and deep learning |
| CN110689919A (en) * | 2019-08-13 | 2020-01-14 | 复旦大学 | Pharmaceutical protein binding rate prediction method and system based on structure and grade classification |
| CN110880354A (en) * | 2019-10-24 | 2020-03-13 | 广东药科大学 | Medicine-target interaction prediction method based on group intelligence |
| CN110957002A (en) * | 2019-12-17 | 2020-04-03 | 电子科技大学 | Drug target interaction relation prediction method based on collaborative matrix decomposition |
| CN111045716A (en) * | 2019-11-04 | 2020-04-21 | 中山大学 | Related patch recommendation method based on heterogeneous data |
| CN111081316A (en) * | 2020-03-25 | 2020-04-28 | 元码基因科技（北京）股份有限公司 | Method and device for screening new coronary pneumonia candidate drugs |
| CN111524546A (en) * | 2020-04-14 | 2020-08-11 | 湖南大学 | Drug-target interaction prediction method based on heterogeneous information |
| CN111540405A (en) * | 2020-04-29 | 2020-08-14 | 新疆大学 | Disease gene prediction method based on rapid network embedding |
| CN111681718A (en) * | 2020-06-11 | 2020-09-18 | 湖南大学 | Medicine relocation method based on deep learning multi-source heterogeneous network |
| CN111785320A (en) * | 2020-06-28 | 2020-10-16 | 西安电子科技大学 | Drug target interaction prediction method based on multilayer network representation learning |
| CN111916145A (en) * | 2020-07-24 | 2020-11-10 | 湖南大学 | Novel coronavirus target prediction and drug discovery method based on graph representation learning |
| CN112070277A (en) * | 2020-08-17 | 2020-12-11 | 杭州电子科技大学 | Hypergraph neural network-based drug-target interaction prediction method |
| CN112133367A (en) * | 2020-08-17 | 2020-12-25 | 中南大学 | Method and device for predicting interaction relation between medicine and target spot |
| CN112216396A (en) * | 2020-10-14 | 2021-01-12 | 复旦大学 | Method for predicting drug-side effect relationship based on graph neural network |
| CN112309509A (en) * | 2019-10-15 | 2021-02-02 | 腾讯科技（深圳）有限公司 | Compound property prediction method, device, computer device and readable storage medium |
| CN112420126A (en) * | 2020-12-07 | 2021-02-26 | 湖南大学 | Drug target prediction method based on multi-source data fusion and network structure disturbance |
| CN112652355A (en) * | 2020-12-08 | 2021-04-13 | 湖南工业大学 | Medicine-target relation prediction method based on deep forest and PU learning |
| CN112863634A (en) * | 2021-01-12 | 2021-05-28 | 山东大学 | Traditional Chinese medicine prescription recommendation method and system based on new crown protein heterogeneous network clustering |
| CN112863693A (en) * | 2021-02-04 | 2021-05-28 | 东北林业大学 | Drug target interaction prediction method based on multi-channel graph convolution network |
| CN112951328A (en) * | 2021-03-03 | 2021-06-11 | 湖南大学 | MiRNA-gene relation prediction method and system based on deep learning heterogeneous information network |
| CN113012770A (en) * | 2021-03-17 | 2021-06-22 | 中南大学 | Medicine-medicine interaction event prediction method, system, terminal and readable storage medium based on multi-modal deep neural network |
| CN113140254A (en) * | 2021-04-28 | 2021-07-20 | 厦门大学 | Meta-learning drug-target interaction prediction system and prediction method |
| CN113270153A (en) * | 2021-05-27 | 2021-08-17 | 南华大学 | Screening method of compound targeting G protein coupled receptor |
| CN113327644A (en) * | 2021-04-09 | 2021-08-31 | 中山大学 | Medicine-target interaction prediction method based on deep embedding learning of graph and sequence |
| CN113611356A (en) * | 2021-07-29 | 2021-11-05 | 湖南大学 | Drug relocation prediction method based on self-supervision graph representation learning |
| CN114121181A (en) * | 2021-11-12 | 2022-03-01 | 东南大学 | Attention mechanism-based traditional Chinese medicine target prediction method for heterogeneous graph neural network |
| CN114242162A (en) * | 2021-07-29 | 2022-03-25 | 北京百度网讯科技有限公司 | Method for establishing drug synergy prediction model, prediction method and corresponding device |
| CN114913938A (en) * | 2022-05-27 | 2022-08-16 | 中南大学 | Small molecule generation method, equipment and medium based on pharmacophore model |
| CN115458045A (en) * | 2022-09-15 | 2022-12-09 | 哈尔滨工业大学 | Drug pair interaction prediction method based on heterogeneous information network and recommendation system |
| WO2023103387A1 (en) * | 2021-12-08 | 2023-06-15 | 深圳大学 | Method and model for predicting transcription factor-target gene interaction |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150324546A1 (en) * | 2012-06-21 | 2015-11-12 | Georgetown University | Method for predicting drug-target interactions and uses for drug repositioning |
| CN106529205A (en) * | 2016-11-03 | 2017-03-22 | 中南大学 | Drug target relation prediction method based on drug substructure and molecule character description information |
| WO2017130078A2 (en) * | 2016-01-27 | 2017-08-03 | King Abdullah University Of Science & Technology | A method to predict and rank new edges in a network |
| CN108520166A (en) * | 2018-03-26 | 2018-09-11 | 中山大学 | A kind of drug targets prediction technique based on multiple similitude network wandering |
-
2019
- 2019-01-15 CN CN201910035394.7A patent/CN109887540A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150324546A1 (en) * | 2012-06-21 | 2015-11-12 | Georgetown University | Method for predicting drug-target interactions and uses for drug repositioning |
| WO2017130078A2 (en) * | 2016-01-27 | 2017-08-03 | King Abdullah University Of Science & Technology | A method to predict and rank new edges in a network |
| CN106529205A (en) * | 2016-11-03 | 2017-03-22 | 中南大学 | Drug target relation prediction method based on drug substructure and molecule character description information |
| CN108520166A (en) * | 2018-03-26 | 2018-09-11 | 中山大学 | A kind of drug targets prediction technique based on multiple similitude network wandering |
Non-Patent Citations (4)
| Title |
|---|
| SEAL, A., AHN, YY. & WILD, D.J.: "Optimizing drug–target interaction prediction based on random walk on heterogeneous networks", 《JOURNAL OF CHEMINFORMATICS》 * |
| X. CHEN, C. C. YAN, X. ZHANG, X. ZHANG, F. DAI, J. YIN, ET AL.: "Drug-target interaction prediction: databases web servers and computational models", 《BRIEFINGS IN BIOINFORMATICS》 * |
| XING CHEN，MING-XI LIU AND GUI-YING YAN: "Drug–target interaction prediction by random walk on the heterogeneous network", 《MOLECULAR BIOSYSTEMS》 * |
| 余蔚明: "药物—靶标相互作用网络预测方法研究", 《中国优秀硕士学位论文全文数据库医药卫生科技辑》 * |
Cited By (54)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110310703A (en) * | 2019-06-25 | 2019-10-08 | 中国人民解放军军事科学院军事医学研究院 | Prediction technique, device and the computer equipment of drug |
| CN110310703B (en) * | 2019-06-25 | 2021-09-07 | 中国人民解放军军事科学院军事医学研究院 | Medicine prediction method and device and computer equipment |
| CN110415763B (en) * | 2019-08-06 | 2023-05-23 | 腾讯科技（深圳）有限公司 | Method, device, equipment and storage medium for predicting interaction between medicine and target |
| CN110415763A (en) * | 2019-08-06 | 2019-11-05 | 腾讯科技（深圳）有限公司 | The interaction prediction method of drug and target, device, equipment and storage medium |
| CN110689919A (en) * | 2019-08-13 | 2020-01-14 | 复旦大学 | Pharmaceutical protein binding rate prediction method and system based on structure and grade classification |
| CN110689919B (en) * | 2019-08-13 | 2023-03-17 | 复旦大学 | Pharmaceutical protein binding rate prediction method and system based on structure and grade classification |
| CN110534153A (en) * | 2019-08-30 | 2019-12-03 | 广州费米子科技有限责任公司 | Target prediction system and method based on deep learning |
| CN110534153B (en) * | 2019-08-30 | 2024-04-19 | 广州费米子科技有限责任公司 | Target prediction system and method based on deep learning |
| CN110598130A (en) * | 2019-09-30 | 2019-12-20 | 重庆邮电大学 | Movie recommendation method integrating heterogeneous information network and deep learning |
| CN110598130B (en) * | 2019-09-30 | 2022-06-24 | 重庆邮电大学 | Movie recommendation method integrating heterogeneous information network and deep learning |
| CN112309509B (en) * | 2019-10-15 | 2021-05-28 | 腾讯科技（深圳）有限公司 | Compound property prediction method, device, computer device and readable storage medium |
| CN112309509A (en) * | 2019-10-15 | 2021-02-02 | 腾讯科技（深圳）有限公司 | Compound property prediction method, device, computer device and readable storage medium |
| CN110880354B (en) * | 2019-10-24 | 2023-06-09 | 广东药科大学 | Drug-target interaction prediction method based on population intelligence |
| CN110880354A (en) * | 2019-10-24 | 2020-03-13 | 广东药科大学 | Medicine-target interaction prediction method based on group intelligence |
| CN111045716A (en) * | 2019-11-04 | 2020-04-21 | 中山大学 | Related patch recommendation method based on heterogeneous data |
| CN111045716B (en) * | 2019-11-04 | 2022-02-22 | 中山大学 | Related patch recommendation method based on heterogeneous data |
| CN110957002A (en) * | 2019-12-17 | 2020-04-03 | 电子科技大学 | Drug target interaction relation prediction method based on collaborative matrix decomposition |
| CN110957002B (en) * | 2019-12-17 | 2023-04-28 | 电子科技大学 | Drug target interaction relation prediction method based on synergistic matrix decomposition |
| CN111081316A (en) * | 2020-03-25 | 2020-04-28 | 元码基因科技（北京）股份有限公司 | Method and device for screening new coronary pneumonia candidate drugs |
| CN111524546B (en) * | 2020-04-14 | 2022-05-03 | 湖南大学 | Drug-target interaction prediction method based on heterogeneous information |
| CN111524546A (en) * | 2020-04-14 | 2020-08-11 | 湖南大学 | Drug-target interaction prediction method based on heterogeneous information |
| CN111540405A (en) * | 2020-04-29 | 2020-08-14 | 新疆大学 | Disease gene prediction method based on rapid network embedding |
| CN111681718A (en) * | 2020-06-11 | 2020-09-18 | 湖南大学 | Medicine relocation method based on deep learning multi-source heterogeneous network |
| CN111785320B (en) * | 2020-06-28 | 2024-02-06 | 西安电子科技大学 | Drug target interaction prediction method based on multi-layer network representation learning |
| CN111785320A (en) * | 2020-06-28 | 2020-10-16 | 西安电子科技大学 | Drug target interaction prediction method based on multilayer network representation learning |
| CN111916145A (en) * | 2020-07-24 | 2020-11-10 | 湖南大学 | Novel coronavirus target prediction and drug discovery method based on graph representation learning |
| CN112070277B (en) * | 2020-08-17 | 2024-04-09 | 杭州电子科技大学 | Medicine-target interaction prediction method based on hypergraph neural network |
| CN112133367A (en) * | 2020-08-17 | 2020-12-25 | 中南大学 | Method and device for predicting interaction relation between medicine and target spot |
| CN112070277A (en) * | 2020-08-17 | 2020-12-11 | 杭州电子科技大学 | Hypergraph neural network-based drug-target interaction prediction method |
| CN112216396A (en) * | 2020-10-14 | 2021-01-12 | 复旦大学 | Method for predicting drug-side effect relationship based on graph neural network |
| CN112216396B (en) * | 2020-10-14 | 2022-11-22 | 复旦大学 | Method for predicting drug-side effect relationship based on graph neural network |
| CN112420126A (en) * | 2020-12-07 | 2021-02-26 | 湖南大学 | Drug target prediction method based on multi-source data fusion and network structure disturbance |
| CN112652355B (en) * | 2020-12-08 | 2023-07-04 | 湖南工业大学 | Drug-target relation prediction method based on deep forest and PU learning |
| CN112652355A (en) * | 2020-12-08 | 2021-04-13 | 湖南工业大学 | Medicine-target relation prediction method based on deep forest and PU learning |
| CN112863634B (en) * | 2021-01-12 | 2022-09-20 | 山东大学 | Traditional Chinese medicine prescription recommendation method and system based on new crown protein heterogeneous network clustering |
| CN112863634A (en) * | 2021-01-12 | 2021-05-28 | 山东大学 | Traditional Chinese medicine prescription recommendation method and system based on new crown protein heterogeneous network clustering |
| CN112863693A (en) * | 2021-02-04 | 2021-05-28 | 东北林业大学 | Drug target interaction prediction method based on multi-channel graph convolution network |
| CN112951328B (en) * | 2021-03-03 | 2022-04-22 | 湖南大学 | MiRNA-gene relation prediction method and system based on deep learning heterogeneous information network |
| CN112951328A (en) * | 2021-03-03 | 2021-06-11 | 湖南大学 | MiRNA-gene relation prediction method and system based on deep learning heterogeneous information network |
| CN113012770B (en) * | 2021-03-17 | 2022-05-10 | 中南大学 | Multi-modal deep neural network based prediction of drug-drug interaction events |
| CN113012770A (en) * | 2021-03-17 | 2021-06-22 | 中南大学 | Medicine-medicine interaction event prediction method, system, terminal and readable storage medium based on multi-modal deep neural network |
| CN113327644A (en) * | 2021-04-09 | 2021-08-31 | 中山大学 | Medicine-target interaction prediction method based on deep embedding learning of graph and sequence |
| CN113327644B (en) * | 2021-04-09 | 2024-05-14 | 中山大学 | Drug-target interaction prediction method based on deep embedding learning of graph and sequence |
| CN113140254A (en) * | 2021-04-28 | 2021-07-20 | 厦门大学 | Meta-learning drug-target interaction prediction system and prediction method |
| CN113140254B (en) * | 2021-04-28 | 2023-08-25 | 厦门大学 | Meta-learning drug-target interaction prediction system and prediction method |
| CN113270153A (en) * | 2021-05-27 | 2021-08-17 | 南华大学 | Screening method of compound targeting G protein coupled receptor |
| CN113611356A (en) * | 2021-07-29 | 2021-11-05 | 湖南大学 | Drug relocation prediction method based on self-supervision graph representation learning |
| CN114242162B (en) * | 2021-07-29 | 2022-11-01 | 北京百度网讯科技有限公司 | Method for establishing drug synergy prediction model, prediction method and corresponding device |
| CN114242162A (en) * | 2021-07-29 | 2022-03-25 | 北京百度网讯科技有限公司 | Method for establishing drug synergy prediction model, prediction method and corresponding device |
| CN114121181A (en) * | 2021-11-12 | 2022-03-01 | 东南大学 | Attention mechanism-based traditional Chinese medicine target prediction method for heterogeneous graph neural network |
| CN114121181B (en) * | 2021-11-12 | 2024-03-29 | 东南大学 | Heterogeneous graph neural network traditional Chinese medicine target prediction method based on attention mechanism |
| WO2023103387A1 (en) * | 2021-12-08 | 2023-06-15 | 深圳大学 | Method and model for predicting transcription factor-target gene interaction |
| CN114913938A (en) * | 2022-05-27 | 2022-08-16 | 中南大学 | Small molecule generation method, equipment and medium based on pharmacophore model |
| CN115458045A (en) * | 2022-09-15 | 2022-12-09 | 哈尔滨工业大学 | Drug pair interaction prediction method based on heterogeneous information network and recommendation system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109887540A (en) | A kind of drug targets interaction prediction method based on heterogeneous network insertion | |
| Wang et al. | Boolean modeling in systems biology: an overview of methodology and applications | |
| CN112364880B (en) | Omics data processing method, device, equipment and medium based on graph neural network | |
| CN103116713B (en) | Based on compound and the prediction of protein-protein interaction method of random forest | |
| CN104798043B (en) | A kind of data processing method and computer system | |
| CN107862173A (en) | A kind of lead compound virtual screening method and device | |
| CN107545151A (en) | A kind of medicine method for relocating based on low-rank matrix filling | |
| US20240055071A1 (en) | Artificial intelligence-based compound processing method and apparatus, device, storage medium, and computer program product | |
| CN113571125A (en) | Drug target interaction prediction method based on multilayer network and graph coding | |
| CN115240762B (en) | Multi-scale small molecule virtual screening method and system | |
| CN111950622A (en) | Behavior prediction method, behavior prediction device, behavior prediction terminal and storage medium based on artificial intelligence | |
| CN113470741A (en) | Drug target relation prediction method and device, computer equipment and storage medium | |
| CN103164631B (en) | A kind of intelligent coordinate expression gene analyser | |
| CN116206688A (en) | Multi-mode information fusion model and method for DTA prediction | |
| Vigneshwari et al. | A study on the application of machine learning algorithms using R | |
| Abdul Raheem et al. | Automating Drug Discovery using Machine Learning | |
| Osman et al. | Hybrid learning algorithm in neural network system for enzyme classification | |
| Mathur | Bioinformatics challenges: a review | |
| CN114999566A (en) | Drug repositioning method and system based on word vector characterization and attention mechanism | |
| CN114678083A (en) | Training method and prediction method of chemical genetic toxicity prediction model | |
| CN112133367A (en) | Method and device for predicting interaction relation between medicine and target spot | |
| CN111383708A (en) | Small molecule target prediction algorithm based on chemical genomics and application thereof | |
| CN110176279A (en) | Lead compound virtual screening method and device based on small sample | |
| McIntyre et al. | Cooperative problem decomposition in Pareto competitive classifier models of coevolution | |
| Lee et al. | A protein interaction verification system based on a neural network algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190614 |