The "Drug-Food Interactions" section of NutriChem-2.0 accepts three types of queries. The first is to query the database using a drug name (or DrugBank ID), which we refer to as a drug-query. The second is to query the database using a plant-based food name (or NCBI Taxonomy ID), which we call a food-query. The third is to query the database using a disease or disease class name (or corresponding IDs in Disease Ontology database), which is referred to as a disease-query. At the query page, the text-input fields have cached frequently used terms for each query. The dropdown box that appears when typing a name will show these terms for user convenience.
The output window has three frames:
The user has the option to vary (a) the minimum support, i.e. the minimum number of jointly targeted proteins in this case, required for an edge (i.e. one drug-food interaction pair) and (b) the maximum number of edges allowed to a node in CWN. By default, we use a minimum of one jointly targeted protein for an edge (can range from 1 to 10) and a maximum of 15 edges for a node (can range from 5 to All edges). The limit on jointly targeted proteins applies to both the CWN and LS, while the limit on number of edges applies only to the CWN. It is, however, not recommended to show all edges in the CWN because this will consume system resources, when massive amounts of data are returned.
The full output list of the search is downloadable using the "Export table" button from the lower-right corner in LS frame and can be opened with Excel. The CWN can be downloaded, using the "Export network" buttons at the lower-left corner, as both PDF-format (static figure) and XGMML-format that can be imported into Cytoscape.
At the top of the interface, there are a few action buttons. The first button "Home" redirects the user to the home page of NutriChem server. Other buttons enable the user to switch among different types of drug-interacting proteins (drug targets, enzymes, transporter and carriers) or all results. Note that the button with “(None)” indicates there are no interactions categorized to that type.
(A) A search with "Imatinib" as the query. The user has the option to query a drug by name or DrugBank ID. (B) By default, NutriChem returns all relevant drug-food interactions, as well as the number of proteins jointly targeted by Imatinib and the phytochemicals. With the buttons at the top, the user can switch to the ones affecting specific drug-interacting protein types (drug targets, enzymes, transporters and carriers) or go back to the Home page and submit a new query. The result page shows the full list of results in the LS frame and 15 drug-foods interactions (by default) in the CWN. The interactions in LS are ordered by the number of jointly targeted proteins. Within the CWN frame, the edge width reflects the number of proteins for each drug-food pair (for which their names are also labelled), whereas the size of food node reflects the total number of interaction pairs between compounds in the involved food and protein targets jointly targeted by drugs (only for drug query and disease query). (C) Clicking on one drug-food pair in LS or one particular edge in CWN will enable the list to display detailed information about the selected drug-food pair, including the mediating protein targets, drug name, types of the drug-interacting proteins, food name and the numbers of plant compounds for each food-target interaction. The user can click on the target name, drug name and food/plant name, after which he/she will be redirected to UniProt, DrugBank and NCBI Taxonomy Browser, respectively. (D) Clicking on the "Number of Compounds" field will show the phytochemicals responsible for the underlying biological activity, and, hence, for the potential drug-food interaction, as well as PubMed evidence for the food-compound association, their ChEMBL compound records, the compound-target bioactivities and the option to export compound structures in SMILES format (via the button at bottom-right corner).
(A) A search with "pomegranate" as the query and “Drug-Food Interactions” selected. The user has the option to query a food by name (may refer to the dropdown box for possible names) or NCBI Taxonomy ID (e.g. “TAXID:22663” for pomegranate). (B) The output page shares the same structure as the “drug query” output page, except that the node sizes in the CWN frame are not differentiated here.
(A) A search with "Diabetes" as the query and “Drug-Food Interactions” selected. The user has the option to query a disease by name (may refer to the dropdown box for possible names), Disease Ontology database ID (e.g. “DOID:9351” for “Diabetes”), or query by simply selecting a disease class name. (B) The result page only shows the drug-food interaction pairs for which the involved drugs are indicated for the query disease or disease class. In LS frame, the different drugs are listed with corresponding drug-food interactions ordered by the number of proteins. For clarity in network visualisation, only the drug with the highest number of interacting foods is shown by default, together with the foods that it potentially interacts with. Users may switch to the network for another drug of interest using the selection box on the top-left corner. (C) Similar to “drug query’, the pairs in the LS frame and the edges in the CWN frame are also clickable to display detailed information. Different fields in the expanded table can be clicked as well for cross-reference to other databases or to show the table of supporting plant compounds.