Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

The Human Kinome Targeted by FDA Approved Multi-Target Drugs and Combination Products: A Comparative Study from the Drug-Target Interaction Network Perspective

  • Ying Hong Li ,

    Contributed equally to this work with: Ying Hong Li, Pan Pan Wang, Xiao Xu Li

    Affiliation Innovative Drug Research and Bioinformatics Group, Innovative Drug Research Centre and School of Pharmaceutical Sciences, Chongqing University, Chongqing, China

  • Pan Pan Wang ,

    Contributed equally to this work with: Ying Hong Li, Pan Pan Wang, Xiao Xu Li

    Affiliation Innovative Drug Research and Bioinformatics Group, Innovative Drug Research Centre and School of Pharmaceutical Sciences, Chongqing University, Chongqing, China

  • Xiao Xu Li ,

    Contributed equally to this work with: Ying Hong Li, Pan Pan Wang, Xiao Xu Li

    Affiliation Innovative Drug Research and Bioinformatics Group, Innovative Drug Research Centre and School of Pharmaceutical Sciences, Chongqing University, Chongqing, China

  • Chun Yan Yu,

    Affiliation Innovative Drug Research and Bioinformatics Group, Innovative Drug Research Centre and School of Pharmaceutical Sciences, Chongqing University, Chongqing, China

  • Hong Yang,

    Affiliation Innovative Drug Research and Bioinformatics Group, Innovative Drug Research Centre and School of Pharmaceutical Sciences, Chongqing University, Chongqing, China

  • Jin Zhou,

    Affiliation Innovative Drug Research and Bioinformatics Group, Innovative Drug Research Centre and School of Pharmaceutical Sciences, Chongqing University, Chongqing, China

  • Wei Wei Xue,

    Affiliation Innovative Drug Research and Bioinformatics Group, Innovative Drug Research Centre and School of Pharmaceutical Sciences, Chongqing University, Chongqing, China

  • Jun Tan,

    Affiliation Institute of Bioinformation, Chongqing University of Posts and Telecommunications, Chongqing, China

  • Feng Zhu

    zhufeng@cqu.edu.cn

    Affiliation Innovative Drug Research and Bioinformatics Group, Innovative Drug Research Centre and School of Pharmaceutical Sciences, Chongqing University, Chongqing, China

Abstract

The human kinome is one of the most productive classes of drug target, and there is emerging necessity for treating complex diseases by means of polypharmacology (multi-target drugs and combination products). However, the advantages of the multi-target drugs and the combination products are still under debate. A comparative analysis between FDA approved multi-target drugs and combination products, targeting the human kinome, was conducted by mapping targets onto the phylogenetic tree of the human kinome. The approach of network medicine illustrating the drug-target interactions was applied to identify popular targets of multi-target drugs and combination products. As identified, the multi-target drugs tended to inhibit target pairs in the human kinome, especially the receptor tyrosine kinase family, while the combination products were able to against targets of distant homology relationship. This finding asked for choosing the combination products as a better solution for designing drugs aiming at targets of distant homology relationship. Moreover, sub-networks of drug-target interactions in specific disease were generated, and mechanisms shared by multi-target drugs and combination products were identified. In conclusion, this study performed an analysis between approved multi-target drugs and combination products against the human kinome, which could assist the discovery of next generation polypharmacology.

Introduction

Intensive efforts in exploring novel targets from the genome of human and other infectious species [13] have discovered hundreds of successful (targeted by approved drugs), hundreds of clinical trial (targeted by clinical trial drugs), and thousands of research targets (targeted by investigational agents only) [4]. Over the past two decades, kinases have become one of the most intensively studied protein classes in the target and drug discovery, with 46 drugs approved by the U.S. Food and Drug Administration (FDA) [46]. However, only a small portion (< 10%) of the human kinome has been established to yield successful targets [5,7]. Distinct from the traditional “one drug one target” philosophy, the polypharmacology refers to a novel paradigm by modulating more than one etiological target [8]. The concept of polypharmacology involves drugs against several targets of disease-related pathways [9] and combination products acting on distinct targets of various physiological responses [10,11]. Among those 46 FDA approved drugs targeting the human kinome, 15 are multi-target drugs, and 14 are approved for use in combination with other drugs [12]. Based on our review of FDA official website [12], imatinib is the first approved multi-target kinase inhibitor, and trastuzumab in combination with paclitaxel are the first approved combination product targeting the human kinome. As shown in Fig 1, a total of 6 multi-target drugs and 7 combination products are approved before 2010, and there is a steady growth after 2010 to 9 multi-target drugs and 13 combination products. Due to the emerging necessities for treating complex diseases by the means of polypharmacology, it is of great interests to identify novel efficacious target pairs from the human kinome [1315].

thumbnail
Fig 1. The approval timeline of FDA approved multi-target drugs and combination products against the human kinome.

https://doi.org/10.1371/journal.pone.0165737.g001

The approved multi-target kinase inhibitors cope with the multifactorial nature of complex diseases by simultaneously aiming at multiple targets [1620]. In the meantime, the approved combination products targeting the human kinome produce synergistic effects triggered by actions converging at a specific pathway site, which enhances its clinical capacity for treating diseases of great complexity [10,21,22]. So far, several studies have been conducted to assess the applicability domain of both multi-target drug and combination product and to analyze their advantages over another [23,24]. The combination products can reduce unwanted compensatory mechanisms [24], achieve high efficacy and selectivity [10], and prevent the drug resistance [25]. However, their effectiveness can be attenuated by the pharmacokinetic variations among individual ingredients and the problem of drug-drug interaction [26]. The multi-target drugs provide an effective approach to avoid the problem of drug–drug interactions [27], but it is very difficult to design an efficacious multi-target drug since the design of a single target drug has already been problematic [23]. So far, the differential advantages of the multi-target drugs and the combination products are still under debate. Therefore, it is in urgent need to conduct a comprehensive comparative study between the multi-target drugs and the combination products.

In this study, the multi-target drugs and the combination products targeting the human kinome approved by FDA were compared. Firstly, their drug-target interaction networks were drawn and analyzed [28], and corresponding targets of great popularity were identified. So far, several works based on the analysis of biological networks have been successfully conducted for prioritizing disease-related microRNA [29], inferring microRNA-disease associations [30,31], and so on. Secondly, the biochemical classes of their aimed target pairs were discussed, and the preferred combinations of biochemical class were discovered. Finally, diseases treated by both multi-target drugs and combination products were identified. Popular efficacious targets of the same disease were also explored. In sum, this study provided a comprehensive comparative analysis between FDA approved multi-target drugs and combination products against the human kinome from the drug-target interaction network perspective, which could help to identify novel target pairs efficacious for the next generation polypharmacology.

Materials and Methods

The proposed strategy used in this study to compare the human kinome targeted by multi-target drugs and combination products from the drug-target interaction network perspective was illustrated in Fig 2. In general, this comparative study could be separated into five steps: 1) collection of approved drugs and combination products from FDA official website; 2) identification of the primary therapeutic target of drugs from the therapeutic targets database; 3) mapping the successful targets to the human kinome; 4) construction of drug-target interaction networks; and 5) network pharmacology analysis based on the phylogenetic tree and biochemical classes of targets.

thumbnail
Fig 2. A framework of proposed comparative strategy used in this study.

https://doi.org/10.1371/journal.pone.0165737.g002

Data collection

All FDA approved multi-target drugs against the human kinome were collected from the FDA website (Drugs@FDA, http://www.accessdata.fda.gov/scripts/cder/drugsatfda/) by following steps. Firstly, 1,767 FDA approved drugs were identified and collected from Drugs@FDA together with their aimed diseases. Secondly, based on the established criterion of primary therapeutic targets provide in S1 Table, the primary therapeutic targets of these drugs were matched from the Therapeutic Target Database (TTD, http://database.idrb.cqu.edu.cn/TTD/) [4], and in total 1,521 approved drugs against 361 primary therapeutic targets were identified. Thirdly, the biochemical classes of protein targets were discovered by mapping their protein name to the UniProt database [32,33]. Drugs against at least one kinase target (EC: 2.7) were identified, which included 46 drugs aiming at 25 targets. Finally, drugs with more than one target were defined as multi-target drugs, and were further confirmed by comprehensive literature search. In total, 15 multi-target drugs against 13 targets in the human kinome were shown in S2 Table.

Combination products approved by FDA against at least one kinase target were also collected from the FDA official website. Firstly, more than 300 combinations were identified, and their disease indications were collected. Secondly, targets of combination products were matched from the TTD [4], and 264 approved combination products against 112 therapeutic targets were identified. Thirdly, the biochemical class of each target was collected from the UniProt database [32,33]. Only the combinations against at least one target of the kinase class were analyzed in this study, which included 20 combination products against 17 targets (S3 Table). 14 targets were from enzyme family with 9 belonging to the kinase class (EC: 2.7), while the remaining 3 were in the families of antigen, structural protein and nucleic sequence.

Identification of the primary therapeutic target of the FDA approved drugs

Primary therapeutic targets of approved multi-target drugs and combination products were identified by a target validation process, which requires three steps of determination [34,35]. 1) whether the aimed target was expressed in the disease-relevant cells or tissues? 2) whether the intended target could be effectively modulated by a drug or drug-like molecule with adequate activity in biochemical assay? 3) whether modulation of target in cell or animal models ameliorated the relevant disease phenotype? In this study, literature search of target validation data in PubMed database [36] was conducted, and three types of validation data were collected, which include experimentally determined potency of drugs against their primary target, observed potency or effects of drugs against disease models linked to their primary target, and the observed effects of target knockout, knockdown, transgenetic, RNA interference, antibody and antisense of the in-vivo model. Only the targets with full target validation information were defined as primary one of the corresponding drug.

Mapping the successful targets to the human kinome

The human kinome containing 556 kinase genes could be represented by a phylogenetic tree [7]. In order to map all successful targets of the studied multi-target drugs and combination products to the tree of human kinome, the gene names of each successful target were identified from the UniProt database [32,33]. As shown in Fig 3, the targets of both multi-target drugs and combination products were highlighted by green circle, while the targets of either multi-target drugs or combination products were highlighted by blue and orange circle respectively.

thumbnail
Fig 3. Targets of approved multi-target drugs and combination products in the phylogenetic tree of the human kinome adapted from Manning and colleagues [7].

The targets of both multi-target drugs and combination products were highlighted by green circle, while the targets of either multi-target drugs or combination products were highlighted by blue and orange circle respectively.

https://doi.org/10.1371/journal.pone.0165737.g003

Construction of drug-target interaction network

The drug-target interaction networks were constructed and displayed using the Cytoscape [37], which was a stand-alone platform for visualizing molecular interactions. 15 multi-target drugs together with their corresponding 13 targets were uploaded to and displayed in the Cytoscape. As shown in Fig 4, multi-target drugs were represented by orange hexagon, and all targets were shown by blue ellipse. In the meantime, 20 combination drugs along with their aiming 17 targets were inputted and shown in the Cytoscape. The network of drug-target interaction was also provided in Fig 5 with the representation of target the same as that in Fig 4. Small molecular drugs in a specific combination product were represented by round rectangle in orange, while the monoclonal antibodies were shown in magenta. The combination products were shown by round rectangles connected by sine-wave. Drug-target interactions were displayed by edges with shapes of arrow for activation and “T” for inhibition.

thumbnail
Fig 4. The drug-target interaction network of the FDA approved multi-target drugs.

Drugs were represented by orange hexagon, and targets were shown by blue ellipse. Drug-target interactions were displayed by edges with shapes of arrow for activation and “T” for inhibition.

https://doi.org/10.1371/journal.pone.0165737.g004

thumbnail
Fig 5. The drug-target interaction network of the FDA approved combination products.

Small molecular drugs in a specific combination product were represented by round rectangle in orange, and monoclonal antibodies were shown by round rectangle in magenta. The combination products were shown by round rectangles connected by sine-wave, and targets were shown by blue ellipse. Drug-target interactions were displayed by edges with shapes of arrow for activation and “T” for inhibition.

https://doi.org/10.1371/journal.pone.0165737.g005

Construction of phylogenetic tree of targets aimed by FDA approved drugs

Sequences of all targets aimed by any of those approved multi-target drugs and combination products were collected from the UniProt database [32,33], and a phylogenetic tree representing the homology distance among those targets was constructed as follows. Firstly, HMMER [38,39] was used to integrate the hidden Markov model for multiple sequence alignment of those studied protein targets. Secondly, the FastTree 2.1.7 [40] was used to construct the approximately-maximum-likelihood phylogenetic tree based on the previous alignment results. For targets with distant homology hardly aligned by HMMER, their distance relationships were constructed based on the nomenclature for enzymes and protein family defined by the Pfam database [39]. Thirdly, the iTOL3.0 [41] was applied to display the tree in Fig 6. The connections shown in the center of the figure represented the target pairs aimed by the multi-target drugs (lines in blue) and the combination products (lines in orange).

thumbnail
Fig 6. Phylogenetic tree representing the homology distance among all targets targeted by approved multi-target drugs and combination products.

The connections shown in the center represented the target pairs aimed by the multi-target drugs (lines in blue) and the combination products (lines in orange). Bars at the end of each connecting line defined the number of multi-target drugs (blue bar) and combination products (orange bar).

https://doi.org/10.1371/journal.pone.0165737.g006

Sub-network of drug-target interaction generated by disease classification

The World Health Organization (WHO) provided the International Classification of Diseases (ICD) as the standard diagnostic tool for epidemiology, health management and clinical purpose. By mapping diseases of all approved multi-target drugs and combination products in this study to the ICD code, sub-networks based on ICD could be generated. Firstly, the multi-target drugs and the combination products of the same disease class (level 2 of ICD-10) were identified, and their corresponding targets were collected. As a result, 11 sub-networks were identified. Secondly, each sub-network was displayed by the Cytoscape [37]. Among them, only 4 were treated by both multi-target drugs and combination products (S1 Fig). The definition of hexagon, ellipse and round rectangle was the same as that in the Figs 4 and 5. All multi-target drugs were highlighted by an orange hexagon line.

Results and Discussion

Successful targets in human kinome of the multi-target drugs and the combination products

As illustrated in Fig 3, only a very small portion (16 out of 556, ~2.9%) of the human kinome was identified as targeted by either the multi-target drugs or the combination products approved by FDA. There were 6 targets (VEGFR2, EGFR, HER2, CDK4, CDK6 and B-Raf) inhibited by both approved multi-target drugs and combination products. The biochemical classes of these 6 targets included the tyrosine kinase class (TK), the tyrosine kinase like class (TKL) and the CDK, MAPK, GSK3 and CLK kinase class (CMGC). In the meantime, 3 targets (JAK3, MEK and PI3K-delta) were aimed only by the combination products, the biochemical classes of which were distinct from each other (the TK class, the homologs of sterile kinase class (STE) and the PIKK class of protein kinases). Moreover, 7 targets (Src, Abl, Ret, Met, PDGFR-alpha, PDGFR-beta and Kit) inhibited only by the multi-target drugs were all in the TK class. Overall, the majority (10 out of 13) of the kinases targeted by the approved multi-target drugs and about half (4 out of 9) of the kinases inhibited by the approved combination products were from the TK class, which revealed it as the most productive class of target in the human kinome.

Drug-target interaction networks of approved multi-target drugs and combination products

The network of the multi-target drugs and their corresponding targets were illustrated in Fig 4. As a frequently used statistical concept in network analysis, degree was applied to assess the interactions between targets and drugs. Degree of a specific node (drug or target) referred to the number of edges (interaction from other nodes) connected to this node. As shown in Fig 4, the maximum degree of the multi-target drugs was 5, and the minimum one was 2. In particular, 1, 4 and 10 multi-target drugs inhibited 5, 3 and 2 primary targets, respectively. Drug of the highest degree was sorafenib, reflecting the complex nature of its approved disease indications (hepatocellular carcinomas, renal cell carcinoma and differentiated thyroid carcinoma).

The maximum degree of targets was 8, and the minimum one was 1. Particularly, 1, 1, 1, 1, 1, 2 and 6 targets were targeted by 8, 7, 5, 4, 3, 2 and 1 multi-target drugs, respectively. Targets of high degree (>5) were VEGFR2 and c-Kit. VEGF and its receptor VEGFR2 were extensively tested primary therapeutic targets currently in clinical for renal cell carcinoma (RCC), and the development of RCC was proved to be relied on VEGF signaling and its receptor [42]. Similar to VEGFR2, c-Kit was proposed essential in the occurrence and development of the metastatic RCC, and its overexpression was well demonstrated in the chromophobe variety of RCC [43]. Thus, VEGFR2 and c-Kit were identified as the most popular therapeutic targets of approved multi-target drugs, and the binding of which could provide substantial influence on the pathogenesis of the RCC.

Fig 5 illustrated the drug-target interaction of all FDA approved combination products together with their corresponding targets. The maximum degree of the combination products was 3, and the minimum one was 2. In particular, 6 and 14 combination products inhibited 3 and 2 primary therapeutic targets, respectively. The combination products of the highest degree (= 3) was used to treat 3 disease indications: 1) HER2-positive breast cancer (trastuzumab plus doxorubicin plus cyclophosphamide plus paclitaxel, trastuzumab plus docetaxel plus carboplatin, trastuzumab plus doxorubicin plus cyclophosphamide plus docetaxel), 2) HER2-positive metastatic gastric or gastroesophageal junction adenocarcinoma (5-FU plus trastuzumab plus cisplatin, trastuzumab plus cisplatin plus capecitabine), and 3) non-small cell lung cancer (necitumumab plus gemcitabine plus cisplatin). Several drugs (trastuzumab, cisplatin and doxorubicin) in those combinations were frequently appearing. Drugs in a combination could maximize their efficacy by synergistic effects, some of which were explicitly reported in previous study [10].

The maximum degree of targets was 4, and the minimum one was 1. Particularly, 2, 4 and 11 targets were aimed by 4, 2 and 1 combination drugs, respectively. Targets of the highest degree were EGFR and human DNA. EGFR enhanced aerobic glycolysis in the triple negative breast cancer (TNBC) to promote tumor growth and immune escape, and targeting its signaling could offer an applicable strategy to treat TNBC [44]. Moreover, cisplatin and doxorubicin were reported to form adduct with human DNA, which in turn resulted in DNA damage of the cancer cell [45]. Thus, EGFR and human DNA were discovered as popular therapeutic targets inhibited by FDA approved combination products.

The phylogenetic tree and the biochemical classes of target pairs aimed by the multi-target drugs and the combination products approved by FDA

As shown in Fig 6, the phylogenetic tree representing the homology distance among targets aimed by those studied multi-target drugs and combination products was constructed. In total, 24 targets were shown. Among those targets, the majority (21 targets) were enzyme. In particular, 16, 3, 1 and 1 targets were from classes of kinase, oxidoreductase, transferase and isomerase, respectively. Clearly, the kinase class was the most popular target class for both multi-target drugs and combination products. 8, 4, 3 and 1 targets were from the family of receptor tyrosine kinase (in green), serine/threonine protein kinase (in blue), non-receptor tyrosine kinase (in earth yellow) and phosphoinositide 3-kinase (in magenta). In the center of Fig 6, lines connecting 2 targets represented target pairs aimed by the multi-target drugs (blue line) and the combination products (orange line), and significant differences could be observed. First of all, all target pairs of the multi-target drugs were within the kinase class with 11 pairs within the family of receptor tyrosine kinase. The rest of the 8 target pairs were within the non-receptor tyrosine kinase or between the receptor tyrosine kinase and (a) the non-receptor tyrosine kinase and (b) the serine/threonine protein kinase. In distinct contrast, target pairs of the combination products were not concentrated in kinase family, and only one target pair was within the family of serine/threonine protein kinase. The rest of the 16 target pairs were between kinase class and other classes such as transferase, structural protein, nucleic sequence, oxidoreductase, isomerase and antigen. In summary, the approved multi-target drugs tended to inhibit target pairs in the human kinome, especially the receptor tyrosine kinase family, while the approved combination products were able to target more diverse pair of targets with distant homology relationship. It is reasonable to conclude that if one wants to design drug to aim at targets of distant homology relationship, the best choice would be the combination product.

Drug-target interaction sub-networks and popular targets of disease classes treated by both multi-target drugs and combination products

Based on the ICD classification, 4 disease classes treated by both multi-target drugs and the combination products were identified. Sub-networks of drug-target interaction within these 4 classes were displayed by the Cytoscape (S1 Fig). Definition of hexagon, ellipse and round rectangle was the same as that in Figs 4 and 5, and the multi-target drugs were highlighted by an orange hexagon line. Those 4 disease classes identified were breast cancer (C50-C50), cancer of respiratory and intrathoracic organs (C30-C39), cancer of digestive organs (C15-C26) and leukemia and lymphoma (C81-C96). All disease classes were in the super ICD class of “neoplasm”, which reflect the complex and multifactorial nature of cancer and the effectiveness of polypharmacology in the treatment of those 4 cancer types. As shown in S1 Fig, targets of some classes were inhibited by both multi-target drugs and combination products, while targets of other classes were aimed by either the multi-target drugs or the combination products.

Breast cancer.

kinases targeted by both multi-target drugs and combination products for treating breast cancer included HER2, EGFR and CDK4/6. (1) the dual targeting of HER2 and EGFR with lapatinib prevented the subsequent activation of disease signaling in a group of high risk breast cancers [46,47]. As an anti-HER2 monoclonal antibody, trastuzumab was frequently in combination with other drugs, like cisplatin, to produce the synergistic effects via the “anti-counteractive” mechanism between the PI3k-AKT pathway and the cell cycle arrest mediated DNA repair [10]. (2) palbociclib blocked the progression of the cancer cell from G1 to S phase by targeting the cyclin D1-CDK4/6 complex [48,49], and letrozole decreased the estrogen effects through inhibiting the aromatase [50]. The combination of palbociclib and letrozole could therefore amplify their drug efficacy on ER-positive breast cancer by simultaneously blocking the cell cycle and the estrogen synthesis of cancer cells [50].

Leukemia and lymphoma.

although no kinase was found to be targeted by both multi-target drugs and combination products for treating leukemia and lymphoma, popular targets of either the multi-target drugs or the combination products were identified. (1) popular targets of the multi-target drugs were ABL, Src and c-Kit. The mutation and translocation of ABL allowed the chronic myelogenous leukemia (CML) to progress [5153]. Although the inhibition of the mutated ABL was efficacious to the CML, a substantial number of patients had failed to respond to the ABL inhibitors by developing drug resistance [52]. Src’s pathway [54] and c-kit’s mutation [55] were reported to contribute significantly to the drug resistance, and were the key adverse prognostic and predictive prognosis factors for leukemia [54,55]. Thus, the multi-target drugs inhibiting ABL together with Src or c-Kit were expected to counteract drug resistance of the existing ABL inhibitors. (2) the chronic lymphocytic leukemia demonstrated significant activation of PI3-kinase as compared to the normal B cells [56]. The only FDA approved combination product (idelalisib plus rituximab) for the treatment of leukemia were developed to inhibit not only the B lymphocytes but also the constitutively activated PI3-kinase [56,57], which therefore amplified the drug efficacy of this combination.

Conclusion

In this study, FDA approved multi-target drugs and combination products targeting the human kinome were compared by the approach of network medicine. Popular targets of the multi-target drugs (VEGFR2 and c-Kit) and the combination products (EGFR and human DNA) were identified. As discussed, the approved multi-target drugs inhibited target pairs in the human kinome, while the combination products tended to target the much more diverse sets of target pairs. Moreover, similar mechanisms shared by approved multi-target drugs and combination products were identified by analyzing the sub-networks of 4 disease classes defined by the ICD. In sum, this study performed an analysis between approved multi-target drugs and combination products against the human kinome, which could help to identify novel target pairs efficacious for the next generation polypharmacology.

Supporting Information

S1 Fig. Sub-networks of drug-target interactions identified by disease classes treated by both multi-target drugs and combination products.

Definition of hexagon, ellipse and round rectangle was the same as that in Figs 4 and 5, and multi-target drugs were highlighted by an orange hexagon line.

https://doi.org/10.1371/journal.pone.0165737.s001

(TIF)

S1 Table. The established criterion of primary therapeutic targets.

https://doi.org/10.1371/journal.pone.0165737.s002

(PDF)

S2 Table. A list of FDA approved multi-target drugs together with their corresponding approval indications and primary therapeutic targets.

https://doi.org/10.1371/journal.pone.0165737.s003

(PDF)

S3 Table. A list of FDA approved combination products together with their corresponding approval indications and primary therapeutic targets.

https://doi.org/10.1371/journal.pone.0165737.s004

(PDF)

Author Contributions

  1. Conceptualization: FZ.
  2. Data curation: YHL HY CYY JT.
  3. Formal analysis: YHL XXL JZ CYY.
  4. Funding acquisition: FZ.
  5. Investigation: YHL PPW XXL.
  6. Methodology: FZ.
  7. Project administration: FZ.
  8. Resources: YHL PPW XXL.
  9. Software: YHL WWX CYY.
  10. Supervision: FZ.
  11. Validation: YHL PPW XXL JT.
  12. Visualization: YHL XXL JZ.
  13. Writing – original draft: FZ.
  14. Writing – review & editing: FZ.

References

  1. 1. Song C, Knopfel T. Optogenetics enlightens neuroscience drug discovery. Nat Rev Drug Discov. 2015;15: 97–109. pmid:26612666
  2. 2. Meng W, Wu Y, He X, Liu C, Gao Q, Ge L, et al. A systems biology approach identifies effective tumor-stroma common targets for oral squamous cell carcinoma. Cancer Res. 2014;74: 2306–2315. pmid:24556718
  3. 3. Flemming A. Target discovery: Blocking BTK in B-cell disorders. Nat Rev Drug Discov. 2010;9: 681. pmid:20811379
  4. 4. Yang H, Qin C, Li YH, Tao L, Zhou J, Yu CY, et al. Therapeutic target database update 2016: enriched resource for bench to clinical drug target and targeted pathway information. Nucleic Acids Res. 2016;44: D1069–1074. pmid:26578601
  5. 5. Rask-Andersen M, Zhang J, Fabbro D, Schioth HB. Advances in kinase targeting: current clinical use and clinical trials. Trends Pharmacol Sci. 2014;35: 604–620. pmid:25312588
  6. 6. Wu P, Nielsen TE, Clausen MH. Small-molecule kinase inhibitors: an analysis of FDA-approved drugs. Drug Discov Today. 2016;21: 5–10. pmid:26210956
  7. 7. Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S. The protein kinase complement of the human genome. Science. 2002;298: 1912–1934. pmid:12471243
  8. 8. Reddy AS, Zhang S. Polypharmacology: drug discovery for the future. Expert Rev Clin Pharmacol. 2013;6: 41–47. pmid:23272792
  9. 9. Terme M, Pernot S, Marcheteau E, Sandoval F, Benhamouda N, Colussi O, et al. VEGFA-VEGFR pathway blockade inhibits tumor-induced regulatory T-cell proliferation in colorectal cancer. Cancer Res. 2013;73: 539–549. pmid:23108136
  10. 10. Jia J, Zhu F, Ma X, Cao Z, Li Y, Chen YZ. Mechanisms of drug combinations: interaction and network perspectives. Nat Rev Drug Discov. 2009;8: 111–128. pmid:19180105
  11. 11. Hopkins AL. Network pharmacology: the next paradigm in drug discovery. Nat Chem Biol. 2008;4: 682–690. pmid:18936753
  12. 12. Drugs@FDA: FDA Approved Drug Products (http://www.accessdata.fda.gov/scripts/cder/drugsatfda/). Offical website of the U.S. Food and Drug Administration. 2016.
  13. 13. Zhu F, Shi Z, Qin C, Tao L, Liu X, Xu F, et al. Therapeutic target database update 2012: a resource for facilitating target-oriented drug discovery. Nucleic Acids Res. 2012;40: D1128–1136. pmid:21948793
  14. 14. Zhu F, Han B, Kumar P, Liu X, Ma X, Wei X, et al. Update of TTD: Therapeutic Target Database. Nucleic Acids Res. 2010;38: D787–791. pmid:19933260
  15. 15. Zhang W, Bai Y, Wang Y, Xiao W. Polypharmacology in drug discovery: a review from systems pharmacology perspective. Curr Pharm Des. 2016.
  16. 16. Dar AC, Das TK, Shokat KM, Cagan RL. Chemical genetic discovery of targets and anti-targets for cancer polypharmacology. Nature. 2012;486: 80–84. pmid:22678283
  17. 17. Puls LN, Eadens M, Messersmith W. Current status of SRC inhibitors in solid tumor malignancies. Oncologist. 2011;16: 566–578. pmid:21521831
  18. 18. Boran AD, Iyengar R. Systems approaches to polypharmacology and drug discovery. Curr Opin Drug Discov Devel. 2010;13: 297–309. pmid:20443163
  19. 19. Xue W, Wang P, Li B, Li Y, Xu X, Yang F, et al. Identification of the inhibitory mechanism of FDA approved selective serotonin reuptake inhibitors: an insight from molecular dynamics simulation study. Phys Chem Chem Phys. 2016;18: 3260–3271. pmid:26745505
  20. 20. Wang P, Yang F, Yang H, Xu X, Liu D, Xue W, et al. Identification of dual active agents targeting 5-HT1A and SERT by combinatorial virtual screening methods. Biomed Mater Eng. 2015;26 Suppl 1: S2233–2239.
  21. 21. Li X, Tong LJ, Ding J, Meng LH. Systematic combination screening reveals synergism between rapamycin and sunitinib against human lung cancer. Cancer Lett. 2014;342: 159–166. pmid:24018642
  22. 22. Tao L, Zhu F, Xu F, Chen Z, Jiang YY, Chen YZ. Co-targeting cancer drug escape pathways confers clinical advantage for multi-target anticancer drugs. Pharmacol Res. 2015;102: 123–131. pmid:26438971
  23. 23. Bolognesi ML. Polypharmacology in a single drug: multitarget drugs. Curr Med Chem. 2013;20: 1639–1645. pmid:23410164
  24. 24. Rosini M. Polypharmacology: the rise of multitarget drugs over combination therapies. Future Med Chem. 2014;6: 485–487. pmid:24649950
  25. 25. Knight ZA, Lin H, Shokat KM. Targeting the cancer kinome through polypharmacology. Nat Rev Cancer. 2010;10: 130–137. pmid:20094047
  26. 26. Srinivas NR. Is there a place for drug combination strategies using clinical pharmacology attributes?—review of current trends in research. Curr Clin Pharmacol. 2009;4: 220–228. pmid:19500074
  27. 27. Cavalli A, Bolognesi ML, Minarini A, Rosini M, Tumiatti V, Recanatini M, et al. Multi-target-directed ligands to combat neurodegenerative diseases. J Med Chem. 2008;51: 347–372. pmid:18181565
  28. 28. Barabasi AL, Gulbahce N, Loscalzo J. Network medicine: a network-based approach to human disease. Nat Rev Genet. 2011;12: 56–68. pmid:21164525
  29. 29. Zeng X, Zhang X, Zou Q. Integrative approaches for predicting microRNA function and prioritizing disease-related microRNA using biological interaction networks. Brief Bioinform. 2016;17: 193–203. pmid:26059461
  30. 30. Liu Y, Zeng X, He Z, Zou Q. Inferring microRNA-disease associations by random walk on a heterogeneous network with multiple data sources. IEEE/ACM Trans Comput Biol Bioinform. 2016.
  31. 31. Zou Q, Li J, Hong Q, Lin Z, Wu Y, Shi H, et al. Prediction of MicroRNA-Disease Associations Based on Social Network Analysis Methods. Biomed Res Int. 2015;2015: 810514. pmid:26273645
  32. 32. Boutet E, Lieberherr D, Tognolli M, Schneider M, Bansal P, Bridge AJ, et al. UniProtKB/Swiss-Prot, the Manually Annotated Section of the UniProt KnowledgeBase: How to Use the Entry View. Methods Mol Biol. 2016;1374: 23–54. pmid:26519399
  33. 33. Bairoch A. The ENZYME database in 2000. Nucleic Acids Res. 2000;28: 304–305. pmid:10592255
  34. 34. Lindsay MA. Target discovery. Nature reviews. 2003;2: 831–838. pmid:14526386
  35. 35. Vidalin O, Muslmani M, Estienne C, Echchakir H, Abina AM. In vivo target validation using gene invalidation, RNA interference and protein functional knockout models: it is the time to combine. Curr Opin Pharmacol. 2009;9: 669–676. pmid:19646923
  36. 36. Sayers EW, Barrett T, Benson DA, Bolton E, Bryant SH, Canese K, et al. Database resources of the National Center for Biotechnology Information. Nucleic Acids Res. 2011;39: D38–51. pmid:21097890
  37. 37. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13: 2498–2504. pmid:14597658
  38. 38. Eddy SR. Profile hidden Markov models. Bioinformatics. 1998;14: 755–763. pmid:9918945
  39. 39. Finn RD, Coggill P, Eberhardt RY, Eddy SR, Mistry J, Mitchell AL, et al. The Pfam protein families database: towards a more sustainable future. Nucleic Acids Res. 2016;44: D279–285. pmid:26673716
  40. 40. Price MN, Dehal PS, Arkin AP. FastTree 2—approximately maximum-likelihood trees for large alignments. PLoS One. 2010;5: e9490. pmid:20224823
  41. 41. Letunic I, Bork P. Interactive Tree Of Life (iTOL): an online tool for phylogenetic tree display and annotation. Bioinformatics. 2007;23: 127–128. pmid:17050570
  42. 42. Rini BI. Vascular endothelial growth factor-targeted therapy in metastatic renal cell carcinoma. Cancer. 2009;115: 2306–2312. pmid:19402073
  43. 43. Marech I, Gadaleta CD, Ranieri G. Possible prognostic and therapeutic significance of c-Kit expression, mast cell count and microvessel density in renal cell carcinoma. Int J Mol Sci. 2014;15: 13060–13076. pmid:25056544
  44. 44. Lim SO, Li CW, Xia W, Lee HH, Chang SS, Shen J, et al. EGFR signaling enhances aerobic glycolysis in triple negative breast cancer cells to promote tumor growth and immune escape. Cancer Res. 2016;76: 1284–1296. pmid:26759242
  45. 45. Stone EL, Citossi F, Singh R, Kaur B, Gaskell M, Farmer PB, et al. Antitumour benzothiazoles. Part 32: DNA adducts and double strand breaks correlate with activity; synthesis of 5F203 hydrogels for local delivery. Bioorg Med Chem. 2015;23: 6891–6899. pmid:26474663
  46. 46. Higa GM, Abraham J. Lapatinib in the treatment of breast cancer. Expert Rev Anticancer Ther. 2007;7: 1183–1192. pmid:17892419
  47. 47. Nelson MH, Dolder CR. Lapatinib: a novel dual tyrosine kinase inhibitor with activity in solid tumors. Ann Pharmacother. 2006;40: 261–269. pmid:16418322
  48. 48. Finn RS, Dering J, Conklin D, Kalous O, Cohen DJ, Desai AJ, et al. PD 0332991, a selective cyclin D kinase 4/6 inhibitor, preferentially inhibits proliferation of luminal estrogen receptor-positive human breast cancer cell lines in vitro. Breast Cancer Res. 2009;11: R77. pmid:19874578
  49. 49. Yamamoto-Ibusuki M, Arnedos M, Andre F. Targeted therapies for ER+/HER2- metastatic breast cancer. BMC Med. 2015;13: 137. pmid:26059247
  50. 50. Yanase T, Suzuki S, Goto K, Nomura M, Okabe T, Takayanagi R, et al. Aromatase in bone: roles of Vitamin D3 and androgens. J Steroid Biochem Mol Biol. 2003;86: 393–397. pmid:14623536
  51. 51. Quintas-Cardama A, Kantarjian H, Cortes J. Flying under the radar: the new wave of BCR-ABL inhibitors. Nat Rev Drug Discov. 2007;6: 834–848. pmid:17853901
  52. 52. Storey S. Chronic myelogenous leukaemia market. Nat Rev Drug Discov. 2009;8: 447. pmid:19483705
  53. 53. Kantarjian H, Jabbour E, Grimley J, Kirkpatrick P. Dasatinib. Nat Rev Drug Discov. 2006;5: 717–718. pmid:17001803
  54. 54. Li S. Src-family kinases in the development and therapy of Philadelphia chromosome-positive chronic myeloid leukemia and acute lymphoblastic leukemia. Leuk Lymphoma. 2008;49: 19–26. pmid:18203007
  55. 55. Wakita S, Yamaguchi H, Miyake K, Mitamura Y, Kosaka F, Dan K, et al. Importance of c-kit mutation detection method sensitivity in prognostic analyses of t(8;21)(q22;q22) acute myeloid leukemia. Leukemia. 2011;25: 1423–1432. pmid:21606963
  56. 56. O'Brien SM, Lamanna N, Kipps TJ, Flinn I, Zelenetz AD, Burger JA, et al. A phase 2 study of idelalisib plus rituximab in treatment-naive older patients with chronic lymphocytic leukemia. Blood. 2015;126: 2686–2694. pmid:26472751
  57. 57. Byrd JC, Woyach JA, Johnson AJ. Translating PI3K-Delta Inhibitors to the Clinic in Chronic Lymphocytic Leukemia: The Story of CAL-101 (GS1101). Am Soc Clin Oncol Educ Book. 2012: 691–694. pmid:24451820