Drug Combination Database Construction And Synergistic Mechanism Analysis

Accumulating scientific and clinical evidences have suggested the use of drug combinations as a safe and effective approach, to treat complicated and refractory diseases. The Drug Combination Database (DCDB) is created and devoted to the research and development of multi-component drugs. By collecting and organizing known examples of drug combinations, it aims to facilitate in-depth analyses of these cases, to summarize patterns of coordinated drug actions, and to provide a basis for theoretical modeling and simulation of such beneficial drug combinations.In DCDB, a drug combination is a tested multi-drug formulation with explicit indications and dose ratio. The objectives of using drug combinations are usually to achieve enhanced patient responses, decreased effective dosage (avoiding high dosage dependent toxicity), reduced or delayed development of drug resistance, and simultaneous enhancement of therapeutic actions and reduction of unwanted actions (efficacy synergism with toxicity antagonism).The drug combinations are manually collected from PubMed literature and the U.S. Food and Drug Administration (FDA) Approved Drug Products with Therapeutic Equivalence Evaluations (Orange Book) and clinicaltrials.gov.178approved multi-component drugs were collected from the latest electronic Orange Book. Another321combinations in different R&D stages were extracted from different references. Among them,40were proven unsuccessful. For each individual drug used in combinations, its chemical, pharmacology and structure information were annotated according to the DrugBank database, PubChem database, MedlinePlus reports, Wikipedia and PubMed. The metabolizing enzymes of the individual drugs were collected from the DrugBank database and literature. For each known molecular target of the drugs used in combinations, its sequence, cellular locations, and functions and structure were curated according to DrugBank, UniProt, PDB and PubMed. The pathway affiliations of each target were curated according to the Pathway Interaction Database (PID), Biocarta pathway database and Reactome.org.For each drug combination, its overall activity is summarized from its FDA label file (if approved) or the article reporting it. Based on whether a drug combination was able to produce the expected improvements over other treatments in clinical trials or pre-clinical studies, each usage of a drug combination (with different dosage or indication) is classified as efficacious, or non-efficacious.A drug combination may have multiple drug interactions between its component drugs, which are the molecular mechanisms underlying its overall effect. These drug interactions are categorized with a two level system. On the first level, drug interactions are divided into pharmacodynamic interactions and pharmacokinetic interactions. Pharmacokinetic interactions are those where one drug can affect the processes by which another drug is absorbed, distributed, metabolized and excreted (ADME). And pharmacodynamic interactions are those where the effects of one drug are changed by the presence of another at its site of action. On the second level, pharmacokinetic interactions are further grouped into four categories:positive or negative regulation of drug transport or permeation, enhanced or reduced drug distribution or localization, drug metabolism interactions, and drug elimination interactions. Similarly, pharmacodynamic interactions are further grouped into five categories based on their "sites of actions". They are:all individual drugs act on the same target, individual drugs act on different targets in the same pathway, individual drugs act on different targets in related pathways, individual drugs act on different targets in cross-talking pathways, and individual drugs act on different targets in pathways with yet unknown relations.Some drug combination petterns have been found through analysising the data in DCDB, but they need further investigation.The current version of DCDB(http://www.cls.zju.edu.cn/dcdb/) collected499drug combinations, involving485individual drugs, from more than6,000references. The volume of DCDB data is expected to grow rapidly with the rising interests from both academia and commercial sectors. Therefore, DCDB will be updated, including update for data content and update with possible changes in database schema and potential integration of new analysis tools. Integrated with the high throughout data set and the high quality molecular interaction network, we could go further in drug combination research.