DrugPattern is a tool used to mine regular rules and patterns behind a list of drugs. As the rapid development of pharmacology techniques such as drug repositioning or drug re-purposing, and network pharmacology, researchers often obtains a list of candidate drugs. However, it is difficult to quantitatively narrow down the drug list. More importantly, it will be very useful if the researcher can find informative rules and patterns behind the list of drugs. DrugPattern is such a tool, which can not only mine regular rules and patterns for a list of drugs but also provide helps for quantitatively re-purposing drugs. For any given drug list, DrugPattern is able to:
Step 1: enter your drug list through left panel.
Step 2: enter your background drug list (format same as step 1).
Note: it is not a necessary choice. By default, DrugPattern uses all drugs included in drug categories as the background.
Step 3: Size of drug category.
Step 4: submit the request.
Drug repositioning or drug re-purposing is the process of finding new indications of known drugs and candidate drugs. Using drug repositioning, pharmaceutical companies can more easily find a "new" drug towards a disease, without a significant number of toxicity tests. A drug-repositioning screen often produces a drug list with a variety of drugs. For the drug list, researchers often select a few candidates for further studies by qualitative expert knowledge. However, no quantitative tools are available to automatically mine regular patterns behind the drugs. Since drugs have many same or similar properties, such as target, pathway and so on, it will be very helpful to find informative rules and patterns behind a list of drugs.
DrugPattern is a web server to mine the information of a list of drugs through enrichment analysis of their properties. DrugPattern integrated drugs into various meaningful categories according to its properties, such as target, pathway, and adverse reaction and so on. Using the DrugPattern, a given drug list will be compared to collected drug sets and the significant properties of the given drug list can be rapidly identified. Then, researchers can better understand the given drug list and obtain useful indication from the result of analysis.
The detailed usage of the DrugPattern is as followings:
 Visiting DrugPattern: Log on DrugPattern at http://www.cuilab.cn/drugpattern, as shown in Figure 1.
Figure 1.The homepage of DrugPattern
 Analysis of drug list. The analysis procedure is as follows:
(1)Click the menu "Start Analysis" (red "1" in Figure 2).
(2)Enter the drug list (red "2" in Figure 2). Example drug list is:
Figure 2. The analysis procedure of DrugPattern
 Interpretation of the analysis result for DrugPattern.
Figure 3. The result of DrugPattern analysis
(1)The results are organized into eight categories, target, pathway, classification (the chemical structure classification of drug), category (the effect category of drug), ATC (anatomical therapeutic chemical classifications), ADReCS (adverse drug reaction classification), structure similarity, and mechanism similarity (Figure 3).
(2)The users can view the matched drugs and unmatched drugs by moving the mouse to the interested drug set (Figure 3). The users can double click the selected drug set to open a window to copy the interested drugs (red “1” in Figure 4).
Figure 4. The window of a drug list in the result of DrugPattern analysis
(3)The users can click on the title of result table to rank the results by values of Fold (The real matched number/expected matched number) (red “1” in Figure 5), P-value (red “2” in Figure 5), Bonferroni (corrected p value by Bonferroni) (red “3” in Figure 5), FDR (corrected p value by FDR) (red “4” in Figure 5). The significantly enriched drug sets are putatively associated with the given drug list, the most significant value means the strongest tendency.
Figure 5. The ranked result of DrugPattern analysis
(4)Draw a barplot for interested drug sets.
1. Select the interested drug sets by clicking on the checkbox before the drug set name(red “1” in Figure 6).
2. Click button “BarPlot” (red “2” in Figure 6). The histogram will be shown in a new window.
Figure 6. Draw a barplot for interested drug sets
3. The result of an example barplot is shown as Figure 7.
Figure 7. A barplot of interested drug sets.
The x-axis shows the names of drug sets and the y-axis represents the significance value, (-1)*log(p-value).
We try to understand life using computing.
Dr. Qinghua Cui
Department of Biomedical Informatics
Peking University Health Science Center
38 Xueyuan Rd
Beijing, China 100191
Citation: Huang C, Yang W, Wang J, Zhou Y, Geng B, Kararigas G, Yang J, and Cui Q. The DrugPattern tool for drug set enrichment analysis and its prediction for beneficial effects of oxLDL on type 2 diabetes. J Genet Genomics. 2018 Jul 24. pii: S1673-8527(18)30122-X.