CN112927766B - Method for screening disease combination drug - Google Patents
Method for screening disease combination drug Download PDFInfo
- Publication number
- CN112927766B CN112927766B CN202110344727.1A CN202110344727A CN112927766B CN 112927766 B CN112927766 B CN 112927766B CN 202110344727 A CN202110344727 A CN 202110344727A CN 112927766 B CN112927766 B CN 112927766B
- Authority
- CN
- China
- Prior art keywords
- drug
- gene target
- target
- score
- candidate
- 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
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/50—Molecular design, e.g. of drugs
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H70/00—ICT specially adapted for the handling or processing of medical references
- G16H70/40—ICT specially adapted for the handling or processing of medical references relating to drugs, e.g. their side effects or intended usage
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medicinal Chemistry (AREA)
- Pharmacology & Pharmacy (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Medical Informatics (AREA)
- Toxicology (AREA)
- Public Health (AREA)
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Crystallography & Structural Chemistry (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Theoretical Computer Science (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention discloses a method for screening disease combination drugs, which comprises the following steps: 1) For a target disease, obtaining a gene target point associated with the target disease to obtain a gene target point set S; constructing a protein interaction relation network according to the set S; 2) For each given drug, determining a candidate drug set M according to the action score of each gene target in the drug, wherein the gene target set corresponding to each candidate drug is T; 3) Calculating the network proximity of each candidate drug and the target disease based on the protein interaction relationship network, the set T and the set S, and selecting N candidate drugs with the highest network proximity; 4) Calculating the correlation coefficient between the gene target sets corresponding to any two candidate drug combinations in the N candidate drugs and the network distance score of the drugs to determine the final score of the corresponding drug combination; selecting a plurality of candidate drug combinations with the highest final scores as the combined drugs of the target diseases.
Description
Technical Field
The invention relates to the technical field of drug screening, in particular to a method for screening single drugs and combined drugs of diseases based on drug and disease targets and a protein-related action relationship network.
Background
The medicine plays a role in human diseases and is established on a certain molecular action mechanism, and a complex protein interaction relation network of human is just reflected by the molecular action mechanism; small molecule drugs generally act on target gene proteins directly or indirectly through one or more targets to achieve therapeutic effects. In the research and development of the traditional combined medicine, the action effects of a plurality of medicine combinations are compared through a high-flux medicine screening test, so that the medicine combination with the synergistic action is screened out, and a large amount of manpower, material resources and time cost are consumed in the development process.
Although there have been some approaches to synergistic drug combination development based on drug targets and disease protein interaction networks: for example, "a method for predicting a synergistic anticancer drug combination and a pharmaceutical composition" (refer to patent document No. CN 105138862A), which is directed to the cancer field, uses the characteristics of known drug combinations to predict whether an unknown combination is effective, and is limited to the cancer field, and requires that the target disease has the known effective drug combination as a training set, so that the method has low universality; for example, "a drug discovery method based on a thermal diffusion network and its application" (refer to patent document No. CN 107451423A), the method enriches sub-networks of diseases and drug targets to obtain drugs related to target diseases, but cannot quantitatively sequence the drugs, and for diseases with hundreds of drugs, it is impossible to provide drugs that are worth developing with higher priority in a smaller range, and for drug research and development, the cost problem of a large-scale drug screening test still cannot be solved; for example, a "method for determining synergy of drug combinations based on a gene network" (refer to patent document No. CN 101751508A), which can score the strength of synergy between drugs, but as described in the specification, the robustness is greatly reduced when drug genes are reduced, and a single drug is not quantitatively evaluated, so that a wide-range screening cannot be performed.
It is therefore valuable to propose a method of drug development: important target genes of the medicines are stably given, the single medicines and the combined medicines are quantitatively scored aiming at target diseases with universality, accuracy and effectiveness, a candidate list of the single medicines and the combined medicines with proper quantity is given, and the cost problem of a large number of test verifications is avoided; and aiming at some developed new drugs, relocation development of the drugs can be carried out.
Disclosure of Invention
Aiming at the defects and shortcomings in the prior art, the invention aims to provide a method for screening disease combination drugs based on a complex network.
The invention can realize the development of single medicine and medicine combination aiming at target diseases at low cost and quickly.
A target disease is given, disease targets are collected, and all small molecule drug targets which are on the market are calculated; screening a single drug based on the score between a target disease target and a marketed drug target through a complex relationship network of human protein interaction, then taking the drug ranked in the front, calculating the score between the combined targets of any two drugs, and screening out a new drug treatment combination of the target disease; the whole steps constitute a single medicine and medicine combination development method aiming at the target diseases.
Step one, collecting data of interaction relation between disease target and human protein
Collecting and integrating gene targets associated with the target disease from all or part of five public databases of GWASCACatlog, disGeNet, malaCards, eDGAR, openTarget and KEGG disease, or finally obtaining a gene target set S of the target disease based on a disease protein target screening method (refer to patent application with publication number CN 111640468A).
Human protein interaction relationship (abbreviated in english as PPI) data, which were collected mainly by literature and database screening. The published PPI of high-throughput Y2H experiments, PPI databases such as HPRD, dbPTM, bioGRID, PINA, HPRD, MINT, intAct, innateDB, instruct and the like, and PPI reported in the literature are mainly used for obtaining all human protein interaction relations to form PPI networks through data integration. After the PPI network is formed, the PPI network is mainly used for one-step target point expansion and calculating the distance between target point sets based on the network.
Step two, calculating drug target
After each small molecule drug which is approved to be listed by currently obtained drug regulatory agencies (such as FDA and NMPA), three appointed public drug databases of drug Bank, chEMBL and STITCH are simultaneously searched (other databases can be selected for calculation, but the three databases selected by the method and the calculation method are verified to be effective), and the drug is directly used as the drugThe gene targets acting indirectly or indirectly are designated as s1, s2, s3. Meanwhile, in the ChEMBL and STITCH databases, the action scores of the drug and the corresponding gene target, denoted as vectors a1 (corresponding to the target in s 2) and a2 (corresponding to the target in s 3), are obtained, expressed as "pChEMBL value" in the ChEMBL database and "combined score" in the STITCH database. The action score vector a1 of the gene target obtained from the ChEMBL database is subjected to dispersion standardization, and the calculation formula is as follows, so that the standardized vector a1 is obtained′. Then, the action score vector a2 of the gene target obtained from the STITCH database is normalized in the same way to obtain a2′。
Taking a union set of all gene target point sets s1, s2 and s3 obtained from the three databases, and then establishing an action score matrix for the gene target points: that is, if a certain gene target point in the set S appears in the target point list S1 obtained from the drug bank database, the target point is marked as 1, otherwise, the target point is marked as 0; if a certain gene target point in the set S appears in a target point set S2 obtained by the ChEMBL database, adopting a1′The action score of the normalized corresponding target in the vector is marked as 0 if not; if a certain gene target point in the set S appears in a target point list S3 obtained by the STITCH database, adopting a2′And (4) the action score of the normalized corresponding target point in the vector is marked as 0 if not. The final action score matrix is shown below.
D in the above matrixi,ciAnd tiRespectively representing the values obtained by standardizing drug Bank, chEMBL and STITCH targets by formula 1, wherein N is the number of the elements of the set S. Summing the values of each row of the matrix to obtain a final action score of a certain target point of the medicine, wherein the obtained score is greater than or equal toTarget at 1.
About 4000 small-molecule drugs approved by regulatory agencies are currently on the market, the safety of the drugs is fully researched, the drugs with more than one target point are selected, a set of about 2200 candidate drugs is obtained and is marked as M, and a set of target points of each drug is marked as Ti,(i∈M)。
Step three, screening single target disease drug
Based on human protein interaction relation and each drug gene target Ti(i belongs to M) and a target disease gene target set S, calculating the Network proximity of all the medicines in the candidate medicine set M and the target disease by adopting a Network proximity calculation method of a document (refer to Network-based prediction of drug combinations published by Feixiong Cheng in Nature-communications journal), sorting according to the sequence order Rank from small to largei(i belongs to M), and calculating a network proximity Score P _ Scorei(i ∈ M). The P Score is taken as the final Score for a single drug and calculated as follows.
The accuracy evaluation of the medicaments found in the step can be realized by obtaining the percentage of the gold standard medicaments contained in the medicaments classified into the first 100 and the first 200 in all the gold standard medicaments, and the calculation formula is as follows, wherein the gold standard medicaments are the medicaments listed on the market or above the stage III of the clinical test. Analysis of gold standard drugs for multiple diseases has shown that drugs with higher scores are more likely to be potentially effective drugs for the target disease. Therefore, when developing the target disease drug, the drugs ranked at the top can be selected for development.
The formula is mainly used for evaluating the predicted result, and the higher the F value is, the more reliable the predicted result is.
Step four, screening the target disease drug combination
Based on a gene Network formed by human protein interaction relation and 100 medicaments before Network proximity sorting obtained in the previous step are recorded as a set N, the Network distance of any two medicament combinations of the 100 medicaments is calculated by adopting a Network distance calculation method in the existing document (refer to Network-based prediction of drug combinations published by Feixiong Cheng in Nature-communications journal), and the total number of 4950 combinations is recorded as C. According to the gene target point set T corresponding to each candidate drugi(i belongs to N) and human protein interaction relationship, and a gene set which is developed by extracting a drug target point in one step is marked as Ei(i belongs to N), and calculating a correlation coefficient R _ Score among all the medicine combination target point sets in the CkAnd (k ∈ C). Respectively scoring the network proximity Score P _ Score of two drugs in the drug combination and the drug combination correlation coefficient-R _ Score (because the drug combination plays a role in synergy usually based on the complementary action of the drugs among each other, and the combination with inverse correlation can avoid the superposition of side effects caused by the same mechanism), calculating the average value to be used as the final Score of the corresponding drug combination, then screening out the combinations less than or equal to 0 according to the network distance of the drug combination, and sequencing all the combinations according to the final scoreskAnd (k. Epsilon. C). A higher score indicates a higher likelihood that the drug combination is effective. Let the gene set of the drug targets of candidate drug i and candidate drug j developed in one step in the kth drug combination be recorded as EiAnd EjThen, the correlation coefficient calculation formula is as follows.
Scorek=average(-R_Scorek+P_Scorei+P_Scorej),i,j∈N,k∈C
Where Cov is the covariance function, var is the variance function, cov (V)i,Vj) Is a vector ViAnd VjCovariance of (1), var (V)i) Is a vector ViOfDifference, var (V)j) Is a vector VjThe variance of (c). Vi,VjThe gene set E to which the target points of the drugs i and j are further expanded by PPI respectivelyiAnd EjVector for 0/1 coding, the coding function is:
drug combinations with antagonistic effects, or potential severe adverse events, were then excluded by drug-drug interaction data (source drug bank database). In vitro, animal or clinical trial screening, the selection can be performed according to the order of scores as priority, for example, 50 or 100 drug combinations with the highest scores are selected for small-scale trial verification, thereby greatly reducing the screening range.
The invention has the technical effects that:
given a target disease, scores for potentially effective individual drugs and combination drugs are quantified universally against any disease with a defined target. An evaluation method is established, and through verification, the accuracy evaluation index F aiming at the target disease is up to 92 percent at most, which is superior to the existing method. And single drugs and combined drugs with high priority can be flexibly determined according to the score sorting, the range of test screening is narrowed, and the cost is reduced.
Drawings
FIG. 1 is a schematic flow chart of the method of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood by those skilled in the art, the present invention will be further described in detail with reference to the accompanying drawings and specific embodiments.
Example one
Taking 9 tumor diseases as an example, the names are shown in the following table, using the drugs with 9 diseases on the market or clinical trials progressing to stage III or above as gold standard data (the data is from ChEMBL database), calculating the scores of the 9 target diseases and all drugs in the set M through the steps one and two in the method, and then calculating the percentage of the gold standard drugs in the drugs ranked at the top 100 and the top 200 according to the ranking of the scores, and the results are shown in the following table.
It can be seen that the gold standard drugs for 9 tumor diseases are covered, the highest proportion is 92% of the ewing sarcoma related drugs, thereby verifying the accuracy of screening the disease related drugs based on ranking. The drugs with the above-mentioned scores ranked in the top 100 are likely to be developed as candidate drugs for each tumor type, except for those that are already on the market and are being clinically tested.
Example two
By taking Dilated Cardiomyopathy (DCM) as a target disease, 62 seed genes are collected from a database, and 193 dilated cardiomyopathy related gene targets can be obtained through the step I.
Scores for dilated cardiomyopathy and all drugs in the set M can be obtained through the second step. Then, 55 drugs related to the dilated cardiomyopathy can be obtained by a literature mining method, and whether the screening result is effective or not is evaluated by comparing whether the difference of the scores of the drugs and the drugs related to the non-dilated cardiomyopathy is remarkable or not. The results show that the scores of the two are different significantly by p =0.004, and the drug screening results obtained by the method can be seen to be statistically effective.
Taking the medicines which are ranked in the first 50 in the step two, combining every two medicines randomly, and obtaining 554 combinations and combination scores through combination screening and filtering in the step three. Clinical trials with an enrolled set of dilated cardiomyopathy cases were then retrieved from the clinical trial database and tested for the presence of the leading single and combination drug combination in the ranking of the individual and combination drugs screened by the method using the standard data for the combination drug above Phase III of the clinical trial, where only 1 data was retrieved, the drug combination Metoprolol (Metoprolol succinate) and Doxazosin (Doxazosin) from clinical trial NCT01798992 (Phase 4).
The scores for the drug Metopriol and dilated cardiomyopathy ranked 9 th, for the drug Doxazosin and dilated cardiomyopathy ranked 3 rd, and for the drug combination ranked 92 th out of 554 combinations. It can be seen that the standard data for the dilated cardiomyopathy single drugs and drug combinations screened in the present method are all located in the front. In addition to this, a large number of other novel individual drugs and drug combinations have been discovered and developed for the treatment of this disease.
The technical solutions of the present invention are clearly and completely described above, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without any creative work belong to the protection scope of the present invention.
Claims (9)
1. A method of screening for a disease combination drug comprising the steps of:
1) For a target disease, acquiring a gene target associated with the target disease to obtain a gene target set S; constructing a protein interaction relation network according to the gene target set S;
2) For each given drug, respectively inquiring Q designated drug databases to obtain a gene target set directly or indirectly acting on the drug; wherein s isqIn order to obtain a gene target point set directly acting or indirectly acting with the medicine by inquiring from a Q-th appointed medicine database, Q = 1-Q; then establishing an action score matrix W according to the obtained gene target point set; then obtaining the final action score of each gene target of the medicine according to the action score matrix W, and if the final action score of the gene target of the medicine is larger than a set threshold value, taking the target as the final target of the medicine; recording a set formed by all candidate medicines as a set M, and recording a set of gene targets corresponding to all the candidate medicines as a set T; the ith drug candidateThe corresponding gene target point set is marked as Ti,i∈M;
3) Protein interaction relation network constructed based on step 1) and gene target set T of candidate drugsiAnd gene target set S, calculating the network proximity of each candidate drug and the target disease, and sequencing the candidate drugs from small to large; then, calculating the network proximity score of the corresponding candidate drug according to the ranking order of the candidate drug; then selecting the first N candidate medicines with the highest network proximity score;
4) Calculating the network interval of any two candidate drug combinations in the N candidate drugs based on the protein interaction relation network constructed in the step 1) and the obtained N candidate drugs, and collecting the obtained result as C; calculating a correlation coefficient between the gene target sets corresponding to each drug combination in the set C; then determining the final score of the corresponding drug combination according to the respective network proximity score and the drug combination correlation coefficient of the two drugs in the drug combination; then, the candidate drug combinations which are less than or equal to the set value are screened and removed according to the network distance of the candidate drug combinations, and then a plurality of candidate drug combinations with the highest final scores are selected as the combined drugs of the target diseases.
2. The method of claim 1, wherein the action score matrix W is established by: obtaining the action score of each given drug and the corresponding gene target and standardizing the action score; if one gene target p in the gene target set S appears in the gene target set SqIn (3), the gene target p and the gene target set sqThe corresponding element Wpq in the action score matrix W is labeled as the normalized action score corresponding to gene target p, otherwise it is labeled 0.
3. The method of claim 1, wherein the correlation coefficient for the kth drug combination The kth drug combination comprises drug candidate i, drug candidate j; according to the interaction relation network of the drug target and the protein, the gene set E which is developed by extracting the drug target of the candidate drug i in one stepiAccording to the interaction relation network of the drug target and the protein, extracting the gene set E of the drug target of the candidate drug j expanded by one stepj,ViIs EiOf the coded vector, VjIs EjCov is a covariance function and Var is a variance function.
4. The method of claim 3, wherein the final Score for the kth drug combination is Scorek=average(-R_Scorek+P_Scorei+P_Scorej) (ii) a Wherein R _ ScorekP _ Score as the target related coefficient of drug combinationsiP _ Score as the network proximity Score of drug candidate ijThe network proximity score for candidate drug j.
6. The method of claim 1, wherein the set value is 0.
7. The method of claim 1, wherein the action score matrix W is established by: q designated drug databases are drug Bank, chEMBL and STITCH, respectively; the drug is obtained by inquiring drug Bank, chEMBL and STITCHThe gene target point sets with direct action or indirect action are respectively marked as s1, s2 and s3; obtaining the action score of the drug and the corresponding gene target in ChEMBL, and marking as a vector a1; obtaining the action score of the drug and the corresponding gene target in the STITCH, and marking as a vector a2; then respectively carrying out dispersion standardization on the vectors a1 and a2 to obtain standardized vectors a1 'and a2'; if a gene target p in the gene target set S appears in S1, the gene target p and the gene target set S1Marking the corresponding element in the action score matrix W as 1, otherwise marking the element as 0; if a gene target p in the gene target set S appears in S2, the gene target p and the gene target set S2The corresponding element in the action score matrix W is marked as a1', otherwise, the element is marked as 0; if a gene target p in the gene target set S appears in S3, the gene target p and the gene target set S2The corresponding element in the action score matrix W is labeled a2', otherwise 0.
8. A server, comprising a memory and a processor, the memory storing a computer program configured to be executed by the processor, the computer program comprising instructions for carrying out the steps of the method according to any one of claims 1 to 7.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110344727.1A CN112927766B (en) | 2021-03-29 | 2021-03-29 | Method for screening disease combination drug |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110344727.1A CN112927766B (en) | 2021-03-29 | 2021-03-29 | Method for screening disease combination drug |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112927766A CN112927766A (en) | 2021-06-08 |
CN112927766B true CN112927766B (en) | 2022-11-01 |
Family
ID=76176715
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110344727.1A Active CN112927766B (en) | 2021-03-29 | 2021-03-29 | Method for screening disease combination drug |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112927766B (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103301475A (en) * | 2005-12-28 | 2013-09-18 | 斯克里普斯研究所 | Natural antisense and non-coding rna transcripts as drug targets |
CN111243659A (en) * | 2018-11-29 | 2020-06-05 | 中国科学院大连化学物理研究所 | Drug interaction prediction method based on drug multidimensional similarity |
Family Cites Families (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090269772A1 (en) * | 2008-04-29 | 2009-10-29 | Andrea Califano | Systems and methods for identifying combinations of compounds of therapeutic interest |
CN101751508B (en) * | 2008-12-08 | 2011-12-07 | 清华大学 | Drug combination synergistic effect determination method based on gene network |
CN105868518A (en) * | 2015-01-20 | 2016-08-17 | 中国科学院青岛生物能源与过程研究所 | Network pharmacology analysis method for chemical components of over-the-counter species in traditional Chinese medicine preparation |
KR101776094B1 (en) * | 2015-02-10 | 2017-09-07 | 가천대학교 산학협력단 | Apparatus and method for assessing effects of drugs based on networks |
CN104965998B (en) * | 2015-05-29 | 2017-09-15 | 华中农业大学 | The screening technique of many target agents and/or drug regimen |
CN105138862B (en) * | 2015-07-31 | 2017-12-26 | 同济大学 | A kind of Synergistic anti-cancer disease drug combination forecasting method and pharmaceutical composition |
CN107451423B (en) * | 2017-09-29 | 2019-09-06 | 武汉百药联科科技有限公司 | A kind of drug discovery method and its application based on thermal diffusion network |
CN108694991B (en) * | 2018-05-14 | 2021-01-01 | 武汉大学中南医院 | Relocatable drug discovery method based on integration of multiple transcriptome datasets and drug target information |
CN108647489B (en) * | 2018-05-15 | 2020-06-30 | 华中农业大学 | Method and system for screening disease drug target and target combination |
CN110246550B (en) * | 2019-06-12 | 2022-12-06 | 西安电子科技大学 | Drug combination prediction method based on drug similarity network data |
CN111640468B (en) * | 2020-05-18 | 2021-08-24 | 天士力国际基因网络药物创新中心有限公司 | Method for screening disease-related protein based on complex network |
CN112201365A (en) * | 2020-11-10 | 2021-01-08 | 南宁市第二人民医院 | Method for analyzing action mechanism of pachyman against glandular cystitis based on network pharmacology |
CN112382347B (en) * | 2020-11-17 | 2022-09-06 | 南京理工大学 | Synergistic anti-cancer drug combination identification method based on molecular fingerprint and multi-target protein |
-
2021
- 2021-03-29 CN CN202110344727.1A patent/CN112927766B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103301475A (en) * | 2005-12-28 | 2013-09-18 | 斯克里普斯研究所 | Natural antisense and non-coding rna transcripts as drug targets |
CN111243659A (en) * | 2018-11-29 | 2020-06-05 | 中国科学院大连化学物理研究所 | Drug interaction prediction method based on drug multidimensional similarity |
Also Published As
Publication number | Publication date |
---|---|
CN112927766A (en) | 2021-06-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Athanasios et al. | Protein-protein interaction (PPI) network: recent advances in drug discovery | |
Lancaster et al. | Improved prediction of drug‐induced Torsades de Pointes through simulations of dynamics and machine learning algorithms | |
Douglass et al. | A community challenge for a pancancer drug mechanism of action inference from perturbational profile data | |
Wehling | Principles of translational science in medicine: From bench to bedside | |
CN112927765B (en) | Method for repositioning medicine | |
Root et al. | Biological mechanism profiling using an annotated compound library | |
Pagnuco et al. | Analysis of genetic association using hierarchical clustering and cluster validation indices | |
CN114203269B (en) | Anticancer traditional Chinese medicine screening method based on machine learning and molecular docking technology | |
Samimi et al. | cDNA microarray-based identification of genes and pathways associated with oxaliplatin resistance | |
CN111933211B (en) | Cancer accurate chemotherapy typing marker screening method, chemotherapy sensitivity molecular typing method and application | |
CN112053742A (en) | Method and device for screening molecular target protein, computer equipment and storage medium | |
CZ20031090A3 (en) | Method for operating a computer system for carrying out discrete substructure analysis | |
CN111613297A (en) | Method, system and device for establishing traditional Chinese medicine action mechanism model based on network pharmacology | |
Mendonca-Neto et al. | A gene selection method based on outliers for breast cancer subtype classification | |
CN112927766B (en) | Method for screening disease combination drug | |
KR102563901B1 (en) | Prediction method for property of pharmaceutical active ingredient | |
JP2017514470A (en) | Methods and materials for identifying and treating mammals having HER2-positive breast cancer | |
Cassotti et al. | Application of the weighted Power-Weakness Ratio (wPWR) as a fusion rule in ligand-based virtual screening | |
CN110867210B (en) | Methods for screening and determining similarity between epileptic-associated proteins | |
Wu et al. | CarbonAI, A Non-Docking Deep learning based small molecule virtual screening platform | |
Qi | Machine-Learning Based Drug and Supplement Design Platform for Chronic Disease Control | |
Habib et al. | TarDict: A RandomForestClassifier based software predicts drug-target interaction using SMILES | |
Stein | Mapping Molecular Changes in Human Neuropsychiatric Disorders to Zebrafish Behavioral Profiles | |
Zhang et al. | Upregulation of Biomarker Limd1 Was Correlated with Immune Infiltration in Doxorubicin-Related Cardiotoxicity | |
Zhang et al. | Research Article Upregulation of Biomarker Limd1 Was Correlated with Immune Infiltration in Doxorubicin-Related Cardiotoxicity |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |