miRSel is resource for miRNA-gene associations developed at the Ludwig-Maximilians-Universität München. miRSel offers the currently largest collection of literature derived miRNA-gene associations. Comprehensive collections of miRNA-gene associations are important for the development of miRNA target prediction tools and the analysis of regulatory networks. miRSel is updated daily and can be queried via microRNA identifiers, gene/protein names, PubMed queries, and gene ontology (GO) terms. To make sure the completeness of the miRNA-gene associations the information from the other existing databases such as TarBase, miRecords, and miR2Disease has been integrated.
miRSel provides a web interface to retrieve information on miRNA-gene pairs stored in the database (Figure 1). The interface enables the AND-combination of different options to restrict query result sets.
1. miRNAs can be selected based on miRNA identifiers and miRNA gene families.
2. Genes can be selected based on gene names, gene symbols, protein names or database identifiers.
3. A PubMed interface enables arbitrary PubMed keyword queries for searching miRSel. miRNA-gene pairs are reported only if found in PubMed abstracts matching the PubMed query.
4. The gene ontology (GO) option restricts the reported miRNA-gene pairs to genes associated with the selected GO-terms.
How to perform a search?
Step 1: The user can select a database (such as miRSel, TarBase, miRecords, miR2Disease, or no restriction as well). Default none is selected.
Step 2: The user can select a species (such as Homo sapiens, Mouse, and Rat). Default none is selected.
Step 3: The user can query the database through miRNA identifiers, gene names, PubMed queries, and Gene ontology (GO) terms. The user can search the database either by exact matching or fuzzy. The fuzzy search functionality allow users to retrieve information without knowing the exact search keyword.
Step 3.1: Once certain miRNA name, gene/protein name, PubMed query and/or gene ontology is received as a query search term, the system searches and displays matched result set identifiers in a tabular format. Then user can choose from a list the identifier(s) of interest for the corresponding miRNA-gene associations documented in the databases along with PubMed abstract references. For example, in case of miRNA (miR-124) as a search key (Figure 2). The system displays the matched hits in different species (if none species is selected in Step 2) (Figure 2). The system retrieves the related miR-124-gene associations documented in the database in a detailed tabular format by clicking ‘search selected option’ button. [By default the system displays the top five matched hits search results in a tabular format].
Step 4: How to analyze the tabular format of miRNA-gene associations:
The tabular format is divided into seven columns and each row presents a unique miRNA-gene pair.
Step 4.1: The first column presents a detail section corresponding to a selected row of miRNA-gene pair. By clicking the “+” link the system displays the PubMed IDs from which the pair has been extracted including database sources in a tree format. In the matching PubMed tree, the user can select a relevant PubMed ID and view the related information such as Database sources (e.g. miRel, TarBase, etc), miRNA names, gene names, relation (e.g. target, co-expressed) (if annotated), taxonomy (e.g. human, mouse, etc). By clicking the following “Links” link option navigates to the NCBI PubMed database. In case of none literature information displays means the selected pair refers to the NCBI not available abstract. The second column contains the numbering of miRNA-gene pair (in case of same gene synonym mapped to many genes the extracted miRNA-gene pair for the same miRNA and literature reference receives the same number).
Step 4.2: The third column contains the miRNA identifier and by clicking the displayed miRNA name navigates to the source from which the identifier was extracted such as miRBase database.
Step 4.3: The fourth, fifth and sixth columns contain the gene full names, Entrez Gene symbol and Entrez Gene ID reference (by clicking it navigates to the Entrez Gene Database).
Step 4.4: The last column contains the database source(s) for a selected miRNA-gene pair of interest. To make sure the comprehensiveness of miRNA-gene pairs the information from other existing databases (e.g. Tarbase, miRecods etc) has been integrated. The system will display all possible database sources for each pair (if none is selected in step: 1).
Step 4.5: The resulting miRNA-gene pairs can be analyzed graphically by clicking the “view as interaction graph” option and export as txt file.
Step 5: Users can include and exclude the result sets searched via step-3 for detail miRNA-gene association pairs (Step 4). By default all search options (such as miRNA ID, gene ID, PubMed ID, Gene ontology) are enabled for AND-combination of mining results. But user can control these options through available check box options followed by ‘Search selected options’ button for each search query results.
Figure 2 shows the systematic workflow of miRSel search by miRNA ID
The miRSel user interface allows to query occurrences, pairs and associations of miRNAs and genes and to restrict the entries in the database using a number of filter criteria (Figure 1):
· The Strictness filter enforces a strict string matching of occurrences against the dictionary entries (i.e. occurrences with special characters not in the dictionary or wrong case are removed) (default selected).
· The single-sentence filter reports only miRNA and gene/protein pairs co-occurring in single sentences as opposed to pairs co-occurring in abstracts (default abstract).
· The relation-type filter restricts matches to a particular type of miRNA-gene association (default none is selected).
· The taxonomy filter aims to enforce organism specificity of the matches. Our organism specific taxonomy dictionary contains synonyms and MeSH vocabularies for all examined organisms as provided by the NCBI taxonomy database. We define organism specific matches as tri-occurrences of a gene name, a miRNA name and an entry of the taxonomy dictionary (default none is selected).
· The gene synonym filter excludes protein/gene synonyms which are assigned to multiple genes/proteins (ambiguous synonyms) in databases (default none is selected).
· The database filter shows the text mining pairs only if they also contained in other databases and/or computational predictions of miRNA gene targets (default none is selected).
A web based graphical user interface to the database
A schematic workflow of miRSel search by miRNA ID: After entering a complete or partial search key (e.g. a miRNA) (A) the user can select a subset of the matching miRNAs (B). Then, corresponding miRNA-target co-occurrences stored in the database are displayed in a tabular format (C). This table enables the navigation to miRNA or gene pages of primary databases (e.g. D=miRBase, E=Entrez Gene, PubMed abstracts that reference particular co-occurrences (F), or to the database sources for which the pair has been integrated (G). Also, details related to each miRNA-target pair e.g. all possible names for a given miRNA or protein in the literature and comparison results of other databases and sequence prediction can be displayed from the table (H). Finally, a miRNA target interaction graph (I) can be displayed that also enables the navigation to miRNA and gene pages (nodes) or PubMed abstracts (edges).
miRSel is developed by: Haroon Naeem, Robert Kueffner, Gergely Csaba and Ralf Zimmer
If you have any comments or suggestions please write an email to email@example.com