Frequently Asked Questions[FAQ]:
1. What information can I find at the database?
Currently, we have protein-protein interaction, protein/gene regulatory and protein-small molecule interaction information stored in the database.
The interaction information is obtained from several manually curated databases and text mining from literature. Manually curated molecular interaction information was obtained from BioGRID
, NCBI Gene
, and STITCH
Pathway related information was obtained from KEGG database
, Pathway Interaction database
Disease information was obtained from PharmGKB
and KEGG database
Gene ontology terms and related information was obtained from Gene Ontology database
and GOA database
2. How can I search the database?
A user can retrieve interaction information of a protein or a small molecule by querying the name in the Molecular Name
search box. Currently, you can query only one molecule name. The query name has to be a name that is already in the database. The system will provide hint automatically when a partial name is typed in. When there is not hint for a particular name, it is not indexed in the database. Please let us know of any missing names and we will be happy to add them.
The user can also limit the type of interactions by choosing other options below the molecular name entry box, as explained below. When an option is not used, the default is to select all possible values. Again, when a name can be typed in a search box, it has to be a name that is already stored in IMID. Hints will be given when partial names are typed in.
- Molecule Type: select one molecular type from the dropdown list. Currently, there are only protein-protein and protein-small molecule types in IMID.
- Interaction Type: select one interaction type from the dropdown list. There are four interaction types in IMID:
- genetic interaction/association: genetic interaction or association between two genes.
- interaction: general interaction. Sometimes, the specific type of interaction is unknown.
- physical interaction: physical interaction identified by the interaction word.
- regulation: regulation identified by the interaction word.
- Interaction Words: the words that are used to describe the particular interaction in the text.
- Probability: for manually annotated interactions, the probability is 1. For interactions extracted by automatic method, the probabilities are given by the method. We used a protein-protein interaction extraction method we developed earlier (Bioinformatics, 25(12), 1536-1542 (2009), Chowdhary et al.) to perform the large scale PPI extraction from PubMed abstracts.
- Pathway: select a pathway name from the dropdown list.
- Disease: type a disease name in the search box.
- Gene Ontology: type a gene ontology name in the search box.
- Species: type a species name in the search box.
3. How can I understand the network?
The network is coded for better visualization and understanding. The details of the coding are described in the Network Legend
4. What are the software involved in generating the network?
The network is generated using Cytoscape web, a web plugin for Cytoscape. ASP.net, jQuery and AJAX are employed in Server Side scripting.
5. How do I cite IMID?
We are preparing the manuscript. The paper will be posted once it is published. You can also cite the website, integrativebiology.org.
6. Is there any browser requirement for running the search functions?
The interface and search functions were tested and run properly on FireFox, Google Chrome, and IE9. Lower versions of IE do not work due to a compatibility issue.
7. Why do I see both names with and without captitalization?
We differentiate names with and without capitalization. Names with all capital letters often refer to human proteins/genes and names with first letter being capital often refer to mouse/rat proteins/genes. When searching for interactions for a particular name, we output interactions for both names with and without capitalization because a user may not be aware of this nomenclature convention. Since the interactions in mouse/rat may not necessarily exist in human, and vice versa, names with different capitalization are separately shown in both the table and the network view.