CN108959843B - Computer screening method of chemical small molecule drug of target RNA - Google Patents
Computer screening method of chemical small molecule drug of target RNA Download PDFInfo
- Publication number
- CN108959843B CN108959843B CN201810573816.1A CN201810573816A CN108959843B CN 108959843 B CN108959843 B CN 108959843B CN 201810573816 A CN201810573816 A CN 201810573816A CN 108959843 B CN108959843 B CN 108959843B
- Authority
- CN
- China
- Prior art keywords
- rna
- small molecule
- small molecules
- screening
- data set
- 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.)
- Active
Links
- 239000000126 substance Substances 0.000 title claims abstract description 76
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000012216 screening Methods 0.000 title claims abstract description 38
- 229940126586 small molecule drug Drugs 0.000 title claims abstract description 22
- 150000003384 small molecules Chemical class 0.000 claims abstract description 100
- 238000012549 training Methods 0.000 claims abstract description 24
- 201000010099 disease Diseases 0.000 claims abstract description 8
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims abstract description 8
- 238000005065 mining Methods 0.000 claims abstract description 6
- 230000002265 prevention Effects 0.000 claims abstract description 6
- 108091032973 (ribonucleotides)n+m Proteins 0.000 claims description 93
- 239000012634 fragment Substances 0.000 claims description 53
- 230000003993 interaction Effects 0.000 claims description 43
- 125000003729 nucleotide group Chemical group 0.000 claims description 27
- 239000002773 nucleotide Substances 0.000 claims description 25
- 108091028043 Nucleic acid sequence Proteins 0.000 claims description 20
- 238000007637 random forest analysis Methods 0.000 claims description 20
- 238000012360 testing method Methods 0.000 claims description 12
- KBPLFHHGFOOTCA-UHFFFAOYSA-N 1-Octanol Chemical compound CCCCCCCCO KBPLFHHGFOOTCA-UHFFFAOYSA-N 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 10
- 108091070501 miRNA Proteins 0.000 claims description 10
- 229910052739 hydrogen Inorganic materials 0.000 claims description 9
- 239000001257 hydrogen Substances 0.000 claims description 9
- 239000002679 microRNA Substances 0.000 claims description 8
- 238000011156 evaluation Methods 0.000 claims description 7
- 238000002790 cross-validation Methods 0.000 claims description 6
- 238000011161 development Methods 0.000 claims description 5
- 238000011160 research Methods 0.000 claims description 5
- 238000010200 validation analysis Methods 0.000 claims description 5
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 5
- 239000000370 acceptor Substances 0.000 claims description 4
- 238000007877 drug screening Methods 0.000 claims description 4
- 150000001875 compounds Chemical class 0.000 claims description 3
- 230000007613 environmental effect Effects 0.000 claims description 3
- 238000013537 high throughput screening Methods 0.000 claims description 3
- 239000007853 buffer solution Substances 0.000 claims description 2
- -1 functional site Chemical group 0.000 claims description 2
- 229910021645 metal ion Inorganic materials 0.000 claims description 2
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 239000002904 solvent Substances 0.000 claims description 2
- 238000000126 in silico method Methods 0.000 claims 6
- 239000003814 drug Substances 0.000 description 19
- 108090000623 proteins and genes Proteins 0.000 description 18
- 229940079593 drug Drugs 0.000 description 16
- 102000004169 proteins and genes Human genes 0.000 description 15
- 230000000875 corresponding effect Effects 0.000 description 10
- 108020004707 nucleic acids Proteins 0.000 description 10
- 102000039446 nucleic acids Human genes 0.000 description 10
- 150000007523 nucleic acids Chemical class 0.000 description 10
- 108020004414 DNA Proteins 0.000 description 8
- REFJWTPEDVJJIY-UHFFFAOYSA-N Quercetin Chemical compound C=1C(O)=CC(O)=C(C(C=2O)=O)C=1OC=2C1=CC=C(O)C(O)=C1 REFJWTPEDVJJIY-UHFFFAOYSA-N 0.000 description 8
- IYRMWMYZSQPJKC-UHFFFAOYSA-N kaempferol Chemical compound C1=CC(O)=CC=C1C1=C(O)C(=O)C2=C(O)C=C(O)C=C2O1 IYRMWMYZSQPJKC-UHFFFAOYSA-N 0.000 description 8
- MWDZOUNAPSSOEL-UHFFFAOYSA-N kaempferol Natural products OC1=C(C(=O)c2cc(O)cc(O)c2O1)c3ccc(O)cc3 MWDZOUNAPSSOEL-UHFFFAOYSA-N 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 238000000547 structure data Methods 0.000 description 6
- 230000008685 targeting Effects 0.000 description 5
- UBSCDKPKWHYZNX-UHFFFAOYSA-N Demethoxycapillarisin Natural products C1=CC(O)=CC=C1OC1=CC(=O)C2=C(O)C=C(O)C=C2O1 UBSCDKPKWHYZNX-UHFFFAOYSA-N 0.000 description 4
- ZVOLCUVKHLEPEV-UHFFFAOYSA-N Quercetagetin Natural products C1=C(O)C(O)=CC=C1C1=C(O)C(=O)C2=C(O)C(O)=C(O)C=C2O1 ZVOLCUVKHLEPEV-UHFFFAOYSA-N 0.000 description 4
- HWTZYBCRDDUBJY-UHFFFAOYSA-N Rhynchosin Natural products C1=C(O)C(O)=CC=C1C1=C(O)C(=O)C2=CC(O)=C(O)C=C2O1 HWTZYBCRDDUBJY-UHFFFAOYSA-N 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 4
- 230000018109 developmental process Effects 0.000 description 4
- 238000009509 drug development Methods 0.000 description 4
- 235000008777 kaempferol Nutrition 0.000 description 4
- UXOUKMQIEVGVLY-UHFFFAOYSA-N morin Natural products OC1=CC(O)=CC(C2=C(C(=O)C3=C(O)C=C(O)C=C3O2)O)=C1 UXOUKMQIEVGVLY-UHFFFAOYSA-N 0.000 description 4
- 235000005875 quercetin Nutrition 0.000 description 4
- 229960001285 quercetin Drugs 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 108091032955 Bacterial small RNA Proteins 0.000 description 3
- 108020004459 Small interfering RNA Proteins 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 3
- 210000004027 cell Anatomy 0.000 description 3
- 238000000205 computational method Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 238000003066 decision tree Methods 0.000 description 3
- 230000002452 interceptive effect Effects 0.000 description 3
- 108091027963 non-coding RNA Proteins 0.000 description 3
- 102000042567 non-coding RNA Human genes 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 239000004055 small Interfering RNA Substances 0.000 description 3
- 108020000948 Antisense Oligonucleotides Proteins 0.000 description 2
- 108091023037 Aptamer Proteins 0.000 description 2
- 208000004930 Fatty Liver Diseases 0.000 description 2
- 206010019708 Hepatic steatosis Diseases 0.000 description 2
- 206010020772 Hypertension Diseases 0.000 description 2
- 241000700159 Rattus Species 0.000 description 2
- 108091046869 Telomeric non-coding RNA Proteins 0.000 description 2
- 239000004098 Tetracycline Substances 0.000 description 2
- 239000000074 antisense oligonucleotide Substances 0.000 description 2
- 238000012230 antisense oligonucleotides Methods 0.000 description 2
- 230000008827 biological function Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 206010012601 diabetes mellitus Diseases 0.000 description 2
- 208000010706 fatty liver disease Diseases 0.000 description 2
- 238000013467 fragmentation Methods 0.000 description 2
- 238000006062 fragmentation reaction Methods 0.000 description 2
- 239000012535 impurity Substances 0.000 description 2
- 108020004999 messenger RNA Proteins 0.000 description 2
- 238000003032 molecular docking Methods 0.000 description 2
- 210000002464 muscle smooth vascular Anatomy 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 231100000240 steatosis hepatitis Toxicity 0.000 description 2
- UCSJYZPVAKXKNQ-HZYVHMACSA-N streptomycin Chemical compound CN[C@H]1[C@H](O)[C@@H](O)[C@H](CO)O[C@H]1O[C@@H]1[C@](C=O)(O)[C@H](C)O[C@H]1O[C@@H]1[C@@H](NC(N)=N)[C@H](O)[C@@H](NC(N)=N)[C@H](O)[C@H]1O UCSJYZPVAKXKNQ-HZYVHMACSA-N 0.000 description 2
- 229930101283 tetracycline Natural products 0.000 description 2
- 229960002180 tetracycline Drugs 0.000 description 2
- 235000019364 tetracycline Nutrition 0.000 description 2
- 150000003522 tetracyclines Chemical class 0.000 description 2
- 241000894006 Bacteria Species 0.000 description 1
- 229930186217 Glycolipid Natural products 0.000 description 1
- 241000282412 Homo Species 0.000 description 1
- 108020005198 Long Noncoding RNA Proteins 0.000 description 1
- 108091007780 MiR-122 Proteins 0.000 description 1
- 238000003559 RNA-seq method Methods 0.000 description 1
- 238000005411 Van der Waals force Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 230000017531 blood circulation Effects 0.000 description 1
- 230000036772 blood pressure Effects 0.000 description 1
- 210000000170 cell membrane Anatomy 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 230000035487 diastolic blood pressure Effects 0.000 description 1
- 230000006806 disease prevention Effects 0.000 description 1
- 239000003596 drug target Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002440 hepatic effect Effects 0.000 description 1
- 208000006454 hepatitis Diseases 0.000 description 1
- 231100000283 hepatitis Toxicity 0.000 description 1
- 238000013090 high-throughput technology Methods 0.000 description 1
- 230000001631 hypertensive effect Effects 0.000 description 1
- 230000008105 immune reaction Effects 0.000 description 1
- 230000009878 intermolecular interaction Effects 0.000 description 1
- 230000009545 invasion Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 125000003473 lipid group Chemical group 0.000 description 1
- 210000004185 liver Anatomy 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 229920002521 macromolecule Polymers 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 208000030159 metabolic disease Diseases 0.000 description 1
- 230000004060 metabolic process Effects 0.000 description 1
- 230000011987 methylation Effects 0.000 description 1
- 238000007069 methylation reaction Methods 0.000 description 1
- 108091051828 miR-122 stem-loop Proteins 0.000 description 1
- 230000004001 molecular interaction Effects 0.000 description 1
- 239000002547 new drug Substances 0.000 description 1
- 230000009437 off-target effect Effects 0.000 description 1
- 230000037361 pathway Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 229960005322 streptomycin Drugs 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000035488 systolic blood pressure Effects 0.000 description 1
- 229940043263 traditional drug Drugs 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 210000004509 vascular smooth muscle cell Anatomy 0.000 description 1
Images
Landscapes
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
- Pharmaceuticals Containing Other Organic And Inorganic Compounds (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention discloses a computer screening method of a chemical small molecule drug of a target RNA, which comprises the following steps: (1) collecting and sorting a training data set, (2) mining the characteristics of a prediction method, (3) creating a prediction method and a model, and (4) verifying the prediction method and the model. The invention can be used for computer screening of chemical small molecules of the target RNA; RNA-based prevention and treatment of major diseases provides new solutions.
Description
Technical Field
The invention relates to a computer screening method of a drug, in particular to a computer screening method of a chemical small molecule drug of a target RNA.
Background
Genes (DNA) are depositors of genetic material, which are responsible for directing the construction of proteins, which are considered molecules that ultimately perform specific biological functions, while RNA is considered an intermediate molecule that links DNA and protein. Therefore, traditionally, much attention has been focused on the study of proteins (including protein-encoding DNA), and little attention has been paid to RNA. Traditional drug development is mainly based on target proteins, for example, more than 95% of drugs recorded in drug bank databases have protein as their target, but most of proteins do not have targetability (drug target), and only about 400 proteins can be targeted until now, so drug development targeting other kinds of molecules is an urgent priority for disease prevention and treatment. In recent years, with the implementation of the human genome project and the ENCODE project, it has been surprisingly found that in humans, DNA capable of encoding proteins accounts for only about 2% of the total DNA, and most of the remaining 98% of DNA is transcribed into RNA but not translated into protein, and is called non-coding RNA (ncRNA). With the rapid development of high-throughput technologies such as RNA-Seq, a large number of non-coding RNAs have been found, for example, in human body, 4 thousands of mirnas (micro RNAs) and 10 thousands of long non-coding RNAs (long non-coding RNAs) have been found. Research shows that the RNA molecules have important biological functions and are closely related to diseases, even messenger RNA has functions which are not limited to communication of DNA and protein, but have various important functions at the RNA level, people begin to realize that RNA is becoming a potential key target of disease intervention, and the research and development of drugs targeting RNA are attracting wide attention.
One large class of RNA-targeting molecules with drug-targeting potential is RNA or DNA (referred to herein as nucleic acids, to distinguish them from "RNA targets"), such as small interfering RNA (siRNA), antisense oligonucleotides (ASO), miRNA, aptamers (aptamers), and the like. For example, the nucleic acid drug Mirvirasen of Roche (Roche) for the treatment of hepatitis c, targeting human liver-specific miRNA miR-122, has begun phase 2 clinical trials. However, nucleic acid drugs naturally have some disadvantages, such as off-target effect (off-target), susceptibility to immune reaction caused by exogenous macromolecules, poor stability, and difficulty in entering cells. These disadvantages, especially the latter two, severely hamper the drugability of nucleic acids. For example, siRNA is degraded after entering blood circulation for as short as a few minutes, and has very poor stability, which is one of the major obstacles for nucleic acid drug development. In addition, after hundreds of millions of years of evolution, in order to resist the invasion of external harmful substances, double-layer lipid cell membranes are evolved, and exogenous nucleic acids are prevented from entering cells, so that target RNA is difficult to regulate, which is another main obstacle of nucleic acid patent drugs. Thus, in addition to continuing intensive research into nucleic acid-based RNA-targeting drugs, the international scientific community has also begun directing eye light to other possible RNA-targeting strategies, where small chemical molecules begin to reveal the headquarters. Chemical small molecules in drug development refer to organic molecules with molecular weights less than 900 daltons.
Chemical small molecules have good stability and are easy to enter cells, the defects of nucleic acid drugs are greatly overcome, and historically, small molecules have been successful in targeting RNA, such as streptomycin and tetracycline (tetracycline) which target RNA of bacteria. However, a major bottleneck which seriously hinders the development of the field at present is the insufficient calculation method for chemical small molecule screening of the target RNA. International groups of topics including applicants have attempted in this field. Such as miRNA-environmental factor (mostly chemical small molecule) bioinformatics database and prediction platform mirenarchitecture based on miRNA transcriptome or mRNA transcriptome, small molecule and miRNA association database SM2miR and prediction algorithm, but the methods essentially predict the association between miRNA and small molecule 'function', and are not true drug prediction targeting miRNA. Although Kuntz laboratories have attempted to apply "protein-small molecule" docking software "Dock 6.0" to "RNA-small molecule" docking, the method has significant drawbacks as follows: 1) it depends on the RNA tertiary structure, but most of the RNA tertiary structure is unknown, and the RNA tertiary structure is different from the protein tertiary structure, the former has poorer rigidity and stronger flexibility; 2) dock 6 is designed for "protein-small molecule" docking, and the physicochemical properties of RNA, force field parameters, and protein are far apart, so Dock 6 cannot be used in RNA. Recently, Disney laboratories first biologically identified some chemical small molecules with bound small RNA fragments of hairpin loops (hairpin) and ridges (bridge), and then, by using the interaction data, they designed the prediction algorithm Informia, but the algorithm is only suitable for small RNA fragments and is not suitable for large RNA molecules, and the latter is more numerous and more complex, and the action mechanism is different from that of small RNA. In addition, because the Informina data, programs, and servers are not disclosed, it is unclear how accurate the Informina data, programs, and servers are. By combining the above analysis, the current preliminary attempts have disadvantages, the problem of screening targeted RNA drugs is still far from the task, and an updated calculation method is needed to supplement the problems.
According to the above analysis, the screening of directly targeted molecular drugs, the spatial structure and force field of the molecules seem to be indispensable, the number of RNA molecules with known spatial structures is few, and the RNA force field is not clear, which seems to be a pair of contradictions that are difficult to reconcile.
Disclosure of Invention
The invention aims to provide a computer screening method of a chemical small molecule drug of a target RNA aiming at the defects of the prior art, the method utilizes RNA sequence source information and chemical small molecule physicochemical properties to construct a random forest model, and can help to screen the chemical small molecule of the target RNA more conveniently and effectively. The chemical small molecules of the present invention refer to organic molecules having a molecular weight of less than 900 daltons.
The purpose of the invention is realized by the following technical scheme:
a computer screening method of a chemical small molecule drug of a target RNA comprises the following steps: (1) collecting and sorting a data set, (2) mining characteristics used for training a prediction method, (3) creating the prediction method and a model, and (4) verifying the prediction method and the model.
Preferably, the step of collecting and collating the data set comprises the steps of:
(a) retrieving and acquiring structures only consisting of RNA and small molecules from a PDB (protein data bank) database, and extracting corresponding information from the structures, wherein the corresponding information comprises the interaction condition of the RNA and the small molecules and the specific interaction position of the RNA and the small molecules, and the information is used as a training data set;
(b) RNA interaction with small molecules outside the PDB database was collected from SMMRNA (Small molecular modules of RNA) databases and literature reports as test data sets.
Preferably, the mining is used for training the features of the prediction method, and comprises the following steps:
(a) extracting related characteristics of RNA including sequence, structure and function;
(b) physicochemical properties of small molecules were calculated, including Number of Hydrogen Bond Acceptors (HBA), Number of Hydrogen Bond Donors (HBD), Octanol/water distribution coefficient (logP), Molar refractive index (MR), Molecular Weight (MW), and Topological Polar Surface Area (TPSA).
Preferably, the relevant features include: nucleotide class, functional site, nucleotide distance and nds (nucleotide distance) curve, nucleotide frequency and pairing status.
The method and the model for creating the prediction comprise the following steps: and (3) adopting a Balanced Random Forest (BRF) model to establish a calculation method for RNA-chemical small molecule interaction prediction.
Since small molecules usually bind only to local regions of RNA, the RNA is first converted into fragments, but the small molecule-bound fragments (positive samples) are much smaller than the unbound fragments (negative samples) within the resulting fragments, and therefore a computational method for creating predictions of RNA-chemical small molecule interactions is used using a Balanced Random Forest (BRF) model.
Preferably, the calculation method for creating the RNA-chemical small molecule interaction prediction by using the balanced random forest model comprises the following steps: and dividing the negative samples in the training data set into a plurality of parts to reduce the quantity difference between each negative sample and each positive sample, respectively matching with the positive samples to perform model training, and summarizing the output results of the models.
The verification prediction method and the model comprise the following steps: and (4) evaluating the performance of the model obtained in the step (3).
Preferably, the performance evaluation comprises: cross validation using a training data set and/or independent validation using a test data set.
Preferably, the performance evaluation comprises: 5 positive and 5 negative predictors were selected for biological validation.
The invention also adopts the following scheme that the chemical small molecule drug computer screening method of the target RNA is applied to a high-throughput screening platform.
The invention also adopts the following scheme that the computer screening method of the chemical small molecule drug of the target RNA is applied to the computer screening by taking the RNA as the target compound.
The invention also adopts the following scheme that the chemical small molecule drug computer screening method of the target RNA is applied to the PDB database.
The invention also adopts the following scheme that the computer screening method of the chemical small molecule drug of the target RNA is applied to the following fields: the application in a high-throughput screening platform; the application in computer screening by taking RNA as a target compound; and/or application in a PDB database; and/or application in SMMRNA databases; and/or in the application of miRNA-based environmental factor development platform mirenenvironment; and/or use in targeted drug screening; and/or in the prevention and treatment of major diseases.
The invention also adopts the following scheme that the chemical small molecule drug computer screening method of the target RNA is applied to the targeted drug selection. By applying the method, the chemical small molecules kaempferol (kaempferol) and Quercetin (Quercetin) of the target lncSHGL are predicted.
The invention also adopts the following scheme that the chemical small molecule drug computer screening method of the target RNA is applied to the prevention and treatment of serious diseases. A new lncRNA, lncSHGL, which plays a key role in the metabolism of hepatic glycolipids and is a new drug target for the intervention of metabolic diseases such as fatty liver, diabetes and the like, is discovered in the early period. By using the method, the combination of kaempferol (kaempferol) and Quercetin (Quercetin) with lncSHGL is predicted, and the two chemical small molecules are potential prevention and treatment medicines for fatty liver and diabetes.
The invention has the beneficial effects that:
aiming at the important problem of chemical small molecule drug screening of RNA which is a novel disease intervention target, the invention creates a calculation method of chemical small molecule screening of target RNA based on machine learning (by using a random forest method) on the basis of analyzing RNA sequence characteristics and small molecule physicochemical properties due to the limitations of few RNA space structure data, flexible structure, unknown force field and the like. The invention can be used for computer screening of chemical small molecules of the target RNA; RNA-based prevention and treatment of major diseases provides new solutions.
The invention provides a new idea, a new strategy and a new method for screening the target RNA medicament.
Description of the drawings:
FIG. 1. nucleotide distances calculated from RNA sequences (sequence is used to predict secondary structure first, and then distance is calculated) are highly correlated with spatial structure calculated nucleotide distances;
FIG. 2.AK098656 has high specificity in vascular smooth muscle cell expression;
FIG. 3 shows that after AK098656 gene transfer, both systolic pressure (a) and diastolic pressure (b) of rats are significantly increased;
fig. 4. results of the computational method cross-validation created (a) and test results on independent SMMRNA and literature-derived independent datasets (b).
Detailed Description
The following examples and experimental examples are intended to illustrate the present invention, but are not intended to limit the scope of the present invention. The present invention will be further described with reference to specific examples and experimental examples.
Example 1:
1. collection and arrangement of RNA-chemical Small molecule interaction data
1) Training data set
And retrieving a structure only consisting of an RNA chain and small molecules in the PDB database, and cleaning the downloaded PDB structure data to be used as a source of a training data set. If all the small molecules contained in the PDB structure are metal ions or solvent molecules in a buffer solution commonly used in structural biology research, or the length of an RNA chain contained in the PDB structure does not exceed 20, the small molecules are not retained. Next, information on RNA-small molecule interactions is extracted from the retained PDB structure. Since 4.0 angstroms (Angstrom) is about the turning point for the weakest hydrogen bonds and the strongest van der Waals forces, 4.0 angstroms is taken as a threshold for judging the interaction between small molecules and RNA. An interaction is considered to exist if the closest distance between the nucleotide and the atom of the small molecule is less than 4.0 angstroms. As the PDB structure as the source of the training data set has fresh RNA-small molecule pairs without interaction, the small molecules involved in all the PDB structures are firstly sorted out to calculate the Euclidean distance between the physicochemical properties of the small molecules, then, the rest small molecules are respectively sequenced according to the Euclidean distance between the rest small molecules and the physicochemical properties of the small molecules contained in the structure according to one or more small molecules interacted with the RNA chain in each PDB structure, and in order to reduce the possibility of generating false negative RNA-small molecule pairs as much as possible, the intersection of the small molecules with the Euclidean distance sequencing between the 80 th quantiles and the 90 th quantiles of the physicochemical properties is selected to be used for artificially generating the RNA-small molecule interaction pairs without interaction.
2) Independent test data set
RNA-small molecule interactions and possible non-interacting RNA-small molecule pairs were collected manually from the literature as test datasets and new RNA-small molecule interaction data not included in the PDB database was obtained from the SMMRNA database.
2. Calculation of RNA-related characteristics and small molecule physicochemical properties
In one aspect, RNA-related features are extracted from a number of sequence, structure and function perspectives, specifically, for each nucleotide, the following features are extracted separately in sequence:
(1) the nucleotide species itself (A, U, C, G and N);
(2) whether a pair is formed with an additional nucleotide;
(3) whether it is the predicted functional site of the Rsite2 algorithm previously proposed by the applicant;
(4) the geometric distance normalized by this nucleotide in secondary structure scores NNDS values:
NNDS＝∑dist(nti-ntj)/∑dist(ntcentroid-ntj)
wherein nti,ntj,ntcentroidThe nucleotide to be detected, any nucleotide in RNA and the coordinate vector of the RNA center are respectively adopted, and the Euclidean distance is adopted when the nucleotide distance is calculated.
Subsequently, as a result of the fragmentation process of the RNA, the above features (1) to (3) are put into the vector of the corresponding fragment, and (4) are converted into an average value to be assigned to the corresponding fragment, whether the fragment interacts with the small molecule is determined according to whether the nucleotide located at the center of the fragment interacts with the small molecule, and the deletion values in the fragments beyond both ends of the RNA sequence are filled with the normalized NDS values of (1) N (2) or (3) or (4) the first or last nucleotide, respectively, by default. Furthermore, the frequency of the individual nucleotides and of the nucleotide triplets is also counted over the individual fragments. The RNA secondary structure used to determine the status of nucleotide pairing results from multiple pathways, including extraction from the PDB structure using RNApdee (http:// rnapdee. cs. put. poznan. pl /), manual annotation according to relevant literature reports and prediction of RNA sequence using RNAfold.
On the other hand, the chemical small molecule Structure files include a Structure Data Format (SDF) file directly obtained from a PDB database and a Simplified Molecular Input Line Entry (SMILES) format file retrieved from a PubChem database (https:// PubChem.ncbi.nlm.nih.gov /) of NCBI. And then, calculating the physicochemical properties of the chemical micromolecule structure file according to the obtained chemical micromolecule structure file by using an Open Babel software package, wherein the physicochemical properties comprise the number of hydrogen bond acceptors HBA, the number of hydrogen bond donors HDA, the octanol/water distribution coefficient MW, the molar refractive index MR, the topological polar surface area TPSA and the like. These indices can be obtained directly as counts or integrated through the physicochemical properties of known small molecule fragments. For example, for a small molecule containing n fragments, the TPSA of each fragment can be queried and calculated by weighted summation of the number of fragments:
3. method for creating RNA-chemical small molecule interaction prediction
1) Calculating RNA-chemical small molecule interaction tendency fraction
Since RNA only interacts locally with small chemical molecules, applicants propose the idea of fragmenting RNA. Therefore, the RNA related characteristics input into the model are obtained based on the RNA sequence fragments, and the model directly predicts whether the RNA sequence fragments interact with the chemical small molecules, and further integrates the prediction result of the fragment level into the RNA molecule level to make an overall assessment on the tendency of the RNA molecule to interact with the chemical small molecules. Therefore, the fragments predicted to have the possibility of interacting with the chemical small molecules in the RNA sequence are firstly found out, the proportion of the fragments comprising the fragments which are predicted to have the possibility of interacting with the chemical small molecules in the RNA sequence and the fragments which are from the left to the right to the most 5 adjacent fragments is calculated, then the fragments are sorted according to the proportion, the ratio of the average value of the proportion of the fragments to the average value of the distance between the central sequences of the fragments is calculated, and the higher the ratio is, the RNA sequence fragments which can act with the chemical small molecules are distributed more densely on the RNA molecule, and the interaction tendency score is taken as a DRIP (Drug-RNA interaction predictor) score.
2) Creating RNA-chemical small molecule interaction prediction models
Because the number of fragments in the data set which do not interact with the small molecules is far more than that of fragments which interact with the small molecules, a Balanced Random Forest (BRF) model which divides the negative samples into a plurality of parts and respectively matches with the positive samples is adopted, and in addition, the number difference between the negative samples and the positive samples in each part is reduced as much as possible, and the negative samples are limited to be divided into 10 parts at most in order to avoid excessively increasing the complexity of the model. The random forest model is constructed by using R-packet randomForest.
A random forest is a phylogenetic classification model (ensemble) which is actually formed by a plurality of decision trees, one decision tree is trained from a part of samples, wherein paths from root nodes to leaf nodes indicate how the value conditions θ (xi) of different features should be combined according to the weight w to realize classification of the selected part of samples. Finally, the random forest model realizes the prediction of the classification vector y by integrating a series of decision trees:
and optimizing in a step-by-step mode in view of more integrated characteristics in the model and adjustable parameters in the construction process. Firstly, because RNA is subjected to fragmentation treatment, the influence of the fragment lengths of different RNA sequences on the model performance is tested; after adjusting the length of the RNA sequence fragment, the characteristic is screened. In a trained random forest model, the importance score of a single feature is expressed as Gini Importance (GI), the classification goodness of the segmentation (split) mode kappa of the feature in each tree is expressed as Gini impurity i (kappa), and then the Gini impurities in all the trees T are summarized to obtain the importance score of the feature population:
testing the influence of different feature combinations on the model performance, wherein the feature combinations comprise all reserved features, each group of RNA related features are respectively removed, and the small molecule physicochemical properties are standardized by using molecular weight and then the molecular weight is reserved or removed; after the characteristic combination is selected, the proportion of positive and negative fragments in a data set is adjusted, the proportion of the positive and negative fragments corresponding to each micromolecule is different, the model prediction result is biased, the proportion of the negative fragments and the positive fragments corresponding to the micromolecules is controlled to the same level by operating the negative fragments which do not interact with the micromolecules, the negative fragments and the positive fragments corresponding to the micromolecules are doubled from 10 to 1 until the proportion is doubled to 640 to 1, for the condition that the quantity of the negative fragments corresponding to the micromolecules is insufficient, gaps are filled by pseudo negative fragments generated by randomly sampling and randomly mutating one nucleotide in the existing negative fragments, the other characteristics of the artificially manufactured pseudo negative fragments except the sequence are kept consistent with the original negative fragments, and for the condition that the quantity of the negative fragments corresponding to the micromolecules is surplus, RNA sequence fragments are clustered inside the negative fragments and between the negative fragments and the positive fragments by using a CD-HIT tool, then preferentially reserving the negative segments similar to the positive segments according to the clustering result, reducing the redundancy inside the negative segments, and ensuring the representativeness of the reserved negative segments as much as possible; then, under the condition of controlling the proportion of positive and negative fragments corresponding to the small molecules, the influence of different RNA sequence lengths on the model performance is compared again; and finally, setting the number of the classification trees in the random forest model to be increased by 100 from 100 to 1000 each time, and comparing and selecting the number of the classification trees.
4. Verification of created RNA-chemical small molecule interaction prediction method
5-fold cross validation is performed on the training data set, and the prediction performance is mainly evaluated by sensitivity (sensitivity), specificity (specificity) and Matthews Correlation Coefficient (MCC), and the evaluation indexes are defined as follows:
since these evaluation indices depend on specific classifier thresholds, we will also plot ROC curves and use the area under the curve AUC values for evaluation in order to fully evaluate the predictor.
The created method is run on a separate test data set to assess its accuracy.
All drug small molecule structure data were downloaded from drug library (https:// www.drugbank.ca /), and models with different parameters set during optimization were applied to the drug library to screen for small molecules that could interact with AK 098656. Each of 5 positive and negative predictions were selected for further biological validation. Because the BIACORE intermolecular interaction analyzer of GE has the advantages of wide applicable sample types (including chemical small molecules and RNA), no need of labeling molecules, real-time property, ultrahigh sensitivity (weak and transient molecular interaction can be monitored), and the like, the BIACORE analyzer of GE is used for verifying the predicted positive and negative results.
Example 2:
1. collection and arrangement of RNA-chemical small molecule interaction data
A set of reliable and proven RNA-chemical small molecule interaction data is the basis for creating a targeted RNA chemical small molecule screening calculation method. To do so, applicants download the relevant data from the PDB database and analyze it for collation as a training data set. In addition, to verify the proposed prediction method, new RNA-chemical small molecule interaction pairs not included in PDB were obtained from SMMRNA (small molecule models of RNA) databases, new experimentally confirmed RNA-chemical small molecule interaction pairs were manually retrieved from published literature, and SMMRNA and literature retrieval results were used together as independent test datasets.
2. Calculation of RNA-related characteristics and small molecule physicochemical Properties
Extracting RNA-related characteristics such as Nucleotide classes, functional sites, Nucleotide Distance and (NDS) curves, Nucleotide frequencies, pairing states and the like from multiple angles such as sequences, structures, functions and the like based on RNA-chemical small molecule interaction data of a training data set; extracting a structure file from chemical small Molecular structure data, and calculating physicochemical properties including the Number of Hydrogen Bond Acceptors (HBA), the Number of Hydrogen Bond Donors (HBD), Octanol/water distribution coefficient (logP), Molar refractive index (MR), Molecular Weight (MW), Topological Polar Surface Area (TPSA), and the like.
3. Creation of RNA-chemical Small molecule interaction prediction method
Because the number of RNA fragments in the data set which do not interact with the chemical small molecules is far more than that of the fragments which interact with the chemical small molecules, a computing method for establishing RNA-chemical small molecule interaction prediction by dividing a negative sample into a plurality of Balanced Random Forest (BRF) models which are respectively matched with a positive sample is adopted. In addition, the optimization is performed in a step-by-step manner in view of the characteristics integrated in the model and the number of adjustable parameters in the construction process.
4. Verification of RNA-chemical small molecule interaction prediction method
In order to verify the accuracy of the created RNA-chemical small molecule interaction prediction method, 5-fold cross validation is carried out on a training data set, and the prediction performance of the random forest model is evaluated by adopting an AUC value. Runs were then made on separate test data sets, also evaluated using AUC values. Finally, the method is applied to the lncRNA-AK098656 which is previously discovered by the applicant and is specific to the vascular smooth muscle, and 5 positive prediction results and 5 negative prediction results are selected for biological verification.
Example 3:
for the research on the computational method of chemical small molecule drug screening of targeted RNA, the applicant has created a miRNA-based environmental factor (mostly chemical small molecules) development platform miREnvironment (Cui et al, bioinformatics 2011). Small interfering proteins are generally functional sites on interfering proteins, and thus determining functional sites of RNA is an important basis for interfering target RNA. The applicant has successively proposed methods for predicting RNA functional sites such as Rsite, Rsite2(Cui et al scientific Reports 2015,2016), SRAMP (Cui et al nucleic Acids Res 2016, m6A methylation site prediction), and PPUS (Cui et al bioinformatics 2015, pseudouracil site prediction). The applicant discloses that functional sites obtained by RNA sequence and spatial structure have significant consistency and correlation (FIG. 1), which indicates that the RNA sequence contains RNA spatial structure information, and further suggests that in the case of extreme lack of RNA spatial structure data and unknown RNA force field, the RNA sequence characteristics can be used for predicting the chemical small molecules interacting with the RNA sequence.
Example 4:
a vascular smooth muscle specific lncRNA-AK098656 (figure 2) was verified to be significantly elevated in the blood of hypertensive patients, and the blood pressure of rats after being transferred with AK098656 gene was significantly elevated (Jin L et al hypertension 2018,71(2): 262-.
Example 5:
applicants have collated more than 300 pairs of RNA-chemical small molecule interactions from PDB database collections. More than 100 pairs of RNA-chemical small molecule interaction pairs were obtained from SMMRNA databases and literature. Analysis shows that some RNA sequence characteristics are related to chemical small molecule interaction, such as triplet frequency, Rsite2 site, etc. and that some small molecule physicochemical properties are related to RNA interaction, such as octanol/water distribution coefficient, topological polar surface area, etc. A prediction method DRIP is preliminarily constructed based on random forests, and 5-fold cross validation results show that the AUC reaches 0.818, and the AUC reaches 0.829 (figure 4) on SMMRNA and literature-derived independent test data sets, so that the created method has certain accuracy in predicting RNA-chemical small molecule interaction.
Claims (8)
1. A computer screening method of chemical small molecule drugs of target RNA is characterized in that: comprises the following steps: (1) collecting and sorting a data set, (2) mining characteristics used for training a prediction method, (3) creating a prediction method and a model, and (4) verifying the prediction method and the model; wherein,
the step (1) of collecting and collating the data sets comprises the steps of:
(a) retrieving and acquiring structures only consisting of RNA and small molecules from a PDB database, and extracting corresponding information from the structures, wherein the corresponding information comprises the interaction condition of the RNA and the small molecules and the specific interaction position of the RNA and the small molecules, and the information is used as a training data set; the training data set is sequentially screened through a first screening condition, a second screening condition and a third screening condition; wherein,
first screening conditions: if all the small molecules contained in the PDB structure are metal ions or solvent molecules in a buffer solution used in structural biology research, or the length of an RNA chain contained in the PDB structure does not exceed 20 nucleotides, the small molecules are not reserved;
second screening conditions: extracting RNA-small molecule interaction information from the PDB structure; adopting 4.0 angstroms as a threshold value for judging the interaction between the small molecules and the RNA; if the nearest distance between the RNA and the atoms of the small molecule is less than 4.0 angstroms, the RNA and the atoms of the small molecule are considered to have interaction, and subsequent operation is carried out;
and (3) third screening conditions: respectively sequencing the small molecules according to Euclidean distances of physicochemical properties between the small molecules and the small molecules contained in the structure according to one or more small molecules which interact with an RNA chain contained in each PDB structure, and selecting an intersection of the small molecules of which the Euclidean distances of the physicochemical properties are 80-90% in a descending order; (b) collecting the interaction data of RNA and small molecules outside the PDB database from an SMMRNA database and literature reports as a test data set;
the step (2) of mining features for training a prediction method comprises the following steps:
(a) extracting RNA sequence fragment related characteristics;
(b) calculating the physicochemical properties of the small molecules;
the step (3) of creating a prediction method and a model comprises the following steps: creating an equalized random forest model configured to obtain RNA sequence segment-related features input to the random forest model and physicochemical property features of small molecules input to the random forest model;
and training the random forest model according to the training data set.
2. The in silico screening method of RNA-targeted chemical small molecule drugs of claim 1, wherein: the step (2) of mining features for training a prediction method comprises the following steps:
(a) the relevant characteristics of the extracted RNA sequence fragment comprise the relevant characteristics of sequence, structure and function;
(b) the physical and chemical properties of the calculated micromolecules comprise the number of hydrogen bond acceptors, the number of hydrogen bond donors, octanol/water distribution coefficients, molar refractive indexes, molecular weights and topological polar surface areas.
3. The in silico screening method of RNA-targeted chemical small molecule drugs according to claim 2, characterized in that: the relevant features include: nucleotide class, functional site, nucleotide distance and NDS curve, nucleotide frequency and pairing status.
4. The in silico screening method of RNA-targeted chemical small molecule drugs of claim 1, wherein: a calculation method for establishing RNA-chemical small molecule interaction prediction by adopting a balanced random forest model comprises the following steps: and dividing the negative samples in the training data set into a plurality of parts to reduce the quantity difference between each negative sample and each positive sample, respectively matching with the positive samples to perform model training, and summarizing the output results of the models.
5. The in silico screening method of RNA-targeted chemical small molecule drugs of claim 1, wherein: the step (4) of verifying the prediction method and the model comprises the following steps: and (4) evaluating the performance of the model obtained in the step (3).
6. The in silico screening method of RNA-targeted chemical small molecule drugs according to claim 5, characterized in that: the performance evaluation comprises the following steps: cross validation using the training data set and/or independent validation using the test data set.
7. The in silico screening method of RNA-targeted chemical small molecule drugs according to claim 5, characterized in that: the performance evaluation comprises the following steps: 5 positive and 5 negative predictors were selected for biological validation.
8. The computer screening method of the RNA-targeted chemical small molecule drug according to claim 1 is applied to the following fields: the application in a high-throughput screening platform; and/or in the application of computer screening by taking RNA as a target compound; and/or application in a PDB database; and/or application in SMMRNA databases; and/or in the application of miRNA-based environmental factor development platform mirenenvironment; and/or use in targeted drug screening; and/or in the prevention and treatment of major diseases.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810573816.1A CN108959843B (en) | 2018-06-06 | 2018-06-06 | Computer screening method of chemical small molecule drug of target RNA |
| PCT/CN2018/090267 WO2019232748A1 (en) | 2018-06-06 | 2018-06-07 | Computer screening method for chemical small molecule medication targeting rna |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810573816.1A CN108959843B (en) | 2018-06-06 | 2018-06-06 | Computer screening method of chemical small molecule drug of target RNA |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108959843A CN108959843A (en) | 2018-12-07 |
| CN108959843B true CN108959843B (en) | 2021-07-06 |
Family
ID=64493024
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810573816.1A Active CN108959843B (en) | 2018-06-06 | 2018-06-06 | Computer screening method of chemical small molecule drug of target RNA |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN108959843B (en) |
| WO (1) | WO2019232748A1 (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102222178A (en) * | 2011-03-31 | 2011-10-19 | 清华大学深圳研究生院 | Method for screening and/or designing medicines aiming at multiple targets |
| CN107075515A (en) * | 2013-11-22 | 2017-08-18 | 米纳治疗有限公司 | C/EBP α compositions and application method |
| CN107058521A (en) * | 2017-03-17 | 2017-08-18 | 中国科学院北京基因组研究所 | A kind of detecting system for detecting human immunity state |
| CN107893078A (en) * | 2017-11-28 | 2018-04-10 | 西安交通大学 | Target siRNA, expression vector and virion and its pharmacy application of synaptotagmin 11 |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101587510A (en) * | 2008-05-23 | 2009-11-25 | 中国科学院上海药物研究所 | Method for predicting compound carcinogenic toxicity based on complex sampling and improvement decision forest algorithm |
| US20100138205A1 (en) * | 2008-10-10 | 2010-06-03 | Los Alamos National Security, Llc | Stochastic molecular binding simulation |
| CN106548196A (en) * | 2016-10-20 | 2017-03-29 | 中国科学院深圳先进技术研究院 | A kind of random forest sampling approach and device for non-equilibrium data |
-
2018
- 2018-06-06 CN CN201810573816.1A patent/CN108959843B/en active Active
- 2018-06-07 WO PCT/CN2018/090267 patent/WO2019232748A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102222178A (en) * | 2011-03-31 | 2011-10-19 | 清华大学深圳研究生院 | Method for screening and/or designing medicines aiming at multiple targets |
| CN107075515A (en) * | 2013-11-22 | 2017-08-18 | 米纳治疗有限公司 | C/EBP α compositions and application method |
| CN107058521A (en) * | 2017-03-17 | 2017-08-18 | 中国科学院北京基因组研究所 | A kind of detecting system for detecting human immunity state |
| CN107893078A (en) * | 2017-11-28 | 2018-04-10 | 西安交通大学 | Target siRNA, expression vector and virion and its pharmacy application of synaptotagmin 11 |
Non-Patent Citations (3)
| Title |
|---|
| Small molecules against RNA targets attract big backers;Asher Mullard等;《Nature Reviews Drug Discovery》;20171128;第16卷;第813-815页 * |
| 基于分子描述符和机器学习方法预测和虚拟筛选乳腺癌靶向蛋白HEC1抑制剂;何冰等;《物理化学学报》;20150930;第31卷(第9期);第1795-1802页 * |
| 基于网络药理学的miRNA和环境因子相互作用分析与建模;崔庆华等;《中国药理通讯》;20121231;第29卷(第3期);第18页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019232748A1 (en) | 2019-12-12 |
| CN108959843A (en) | 2018-12-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Do et al. | Using extreme gradient boosting to identify origin of replication in Saccharomyces cerevisiae via hybrid features | |
| CN106599615B (en) | A kind of sequence signature analysis method for predicting miRNA target gene | |
| WO2016201564A1 (en) | Neural network architectures for linking biological sequence variants based on molecular phenotype, and systems and methods therefor | |
| CN112951327B (en) | Drug sensitivity prediction method, electronic device, and computer-readable storage medium | |
| CN110111840B (en) | Somatic mutation detection method | |
| WO2013190085A1 (en) | Systems and methods for generating biomarker signatures with integrated dual ensemble and generalized simulated annealing techniques | |
| CN111370073B (en) | Medicine interaction rule prediction method based on deep learning | |
| CN113488104B (en) | Cancer driving gene prediction method and system based on local and global network centrality analysis | |
| CN107679367B (en) | Method and system for identifying co-regulation network function module based on network node association degree | |
| WO2023197718A1 (en) | Circular rna ires prediction method | |
| CN112270958A (en) | Prediction method based on hierarchical deep learning miRNA-lncRNA interaction relation | |
| Yones et al. | High precision in microRNA prediction: A novel genome-wide approach with convolutional deep residual networks | |
| CN108959843B (en) | Computer screening method of chemical small molecule drug of target RNA | |
| CN114388063B (en) | Non-differential gene associated with malignant phenotype of tumor cell and screening method and application thereof | |
| Hwang et al. | Big data and deep learning for RNA biology | |
| CN112992273A (en) | Early colorectal cancer risk prediction evaluation model and system | |
| CN113921084B (en) | Multi-dimensional target prediction method and system for disease-related non-coding RNA (ribonucleic acid) regulation and control axis | |
| CN111785319A (en) | Drug relocation method based on differential expression data | |
| KR102376212B1 (en) | Gene expression marker screening method using neural network based on gene selection algorithm | |
| Kuznetsov | Mathematical modeling of avidity distribution and estimating general binding properties of transcription factors from genome-wide binding profiles | |
| Nugraha et al. | Performance analysis of relief and mRMR algorithm combination for selecting features in lupus Genome-Wide Association Study | |
| Wang et al. | Deep Learning Integration with Phenotypic Similarities and Heterogeneous Networks for Drug-Target Interaction Prediction | |
| Uthayopas et al. | PRIMITI: a computational approach for accurate prediction of miRNA-target mRNA interaction | |
| Cheng et al. | Raw signal segmentation for estimating RNA modifications and structures from Nanopore direct RNA sequencing data | |
| CN112820347B (en) | Disease gene prediction method based on multiple protein network pulse dynamics process |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20201027 Address after: Fc108-05, basement 1, building 1, yard 13, Dazhongsi, Haidian District, Beijing 100098 Applicant after: Beijing Jianmu Technology Co., Ltd Address before: 100191 Peking University Health Science Center, Haidian District, Xueyuan Road, 38, Beijing Applicant before: Peking University |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |