|
NIF Beta Test Site |
 |
NIF HELPSample searches for the NIF:
Try these searches using the simple interface by entering the terms as they are written, including quotes. Tab through the different NIF information sources to see the results. Then try them again using the advanced search interface and NIF vocabulary functions. Video tutorials for some of these searches using the advanced search functions may be accessed by clicking on the links. TUTORIALS FOR ADVANCED SEARCHES
NIF FAQ
The current NIF is still in development and is not fully populated. The current NIF literature archive was set up as a prototype, and consists of 15,000 papers from a selected list of neuroscience-related journals. We expect that the corpus of literature will increase to 67,000 articles from an expanded list of journals by the end of February. List of journals archived in prototype: BMC Neurosci, Nat Rev Neurosci. J Neurosci, Int Rev Neurobiol, Nat Neurosci, Int J Neurosci, Neuron, Neuroscience, Neurobiology ( Bp ). List of journals to be added in expanded corpus: Exp Brain Res, Trends Neurosci, J Comp Neurol, Cereb Cortex, Curr Opin Neurobiol, Neurobiology, Annu Rev Neurosci, Brain Research, J Neurophsyiol, Neuroscience. Currently, there are approximately 375 resources registered. Commerical sites are not listed. NIF does not provide direct access to the full articles indexed in the NIF literature archive. The NIF provides access to the relevant articles in both Pub Med and Google Scholar. If the journal article is open access or you have an institutional or personal subscription, you will be able to retrieve the full text. No; the NIF lists only non-profit entities in the NIF registry. Some commercial sites are available through the NIF Web Sites, however. The NIF offers several search options, and it may be that by expanding or refining a search (see "Searching the NIF"), the resource will be returned. However, the NIF is not fully populated and we know that resources may be missing. You can include your favorite resource by contacting Dr. Maryann Martone or by registering the resource through http://neurogateway.org. Registration of resources requires an account that can be requested through the same site. If you would like to see a resource added to the NIF registry, please contact Dr. Maryann Martone. You can also register the resource yourself through http://neurogateway.org. Registration of resources requires an account that can be requested through the same site.
The NIF registry is a database of neuroscience relevant resources that has been manually annotated using a minimal controlled term set to characterize the type of resource (e.g., reagent repository, software tools, disease-specific knowledgebase, etc.) and its biological scope (e.g., neurophysiology, Parkinson's disease, neurodevelopment, etc.). Additional information like contact information, host resource etc. has also been provided. In general, the detailed content within the resource is not characterized in this description. For example, if a site has MRI data on many different neurodegenerative diseases, the site will be classified under "neurodegenerative disease" and not all the specific types of neurodegenerative diseases it contains. Via the NIF Web feature, additional resource content is indexed which can provide more nuanced querying depending on the information presented on that resource's web pages. Content presented on dynamically generated pages, i.e., contained in a database, will not have been indexed in NIF Web and, therefore, is inaccessible to NIF Web queries. The NIF provides access to this deep hidden content through the NIF data federation. Once a resource is registered to the NIF data federation, direct queries can be made against the database through the NIF interface. The NIF data federation currently consists of 5 resources registered at level 3. Registration at level 3 via the NIF data federation tools enables NIF to issue a highly-tuned concept-based query against the data in the resource. Registration at level 3 does not require any re-engineering of the original resource; rather, the NIF tools create a query mediation wrapper for the resource. The resources mapped into this prototype NIF data federation during the current phase of development were primarily used as a quality assurance testing of the NIF concept mapping and query mediation infrastructure. It is anticipated this federation will expand as the NIF project grows. The NIF basic vocabularies are derived from the NIF terminology workshops run by Dr. Daniel Gardner. Experts in various neuroscience domains contributed the major high-level terms that were then arranged in a shallow hierarchy, coded in XML. This coarsely specified set of controlled terms was assembled primarily to ensure the concepts used to index neuroscience resources in the NIF Registry reflected the broadest possible interests across the neuroscience community. To support fine-grained nuanced concept-driven level 3 queries, the NIF basic vocabulary was incorporated into a formal ontology (NIFSTD) constructed according to biomedical ontology best practices (OBO Foundry Principles supported by the National Center for Biomedical Ontologies), which include the re-use of relevant existing ontologies. NIFSTD re-uses the core structure of the BIRNLex, a lexicon of concepts largely covering neuroanatomy across scales, organism taxonomy, and disease and incorporates community ontologies such as the Gene Ontology and Neuronames. BIRNLex is encoded in OWL, the web ontology language. The NIF vocabularies may be viewed and searched through the simple ontology browser available here. Use of the browser requires Java Webstart. Through the OntologEZ browser, users can search for a term, navigate the NIF hierarchy and view detailed properties of each entity. When using the browser, please make sure the "Search All Classes" is checked, before issuing a query. Currently, the NIF vocabularies are curated and maintained by the NIF. If you would like to add terms to the vocabulary, you may do so here. If you would like to suggest additional content in the meanwhile, please contact Bill Bug. The concept-based search probes data sets based on shared meaning as opposed to terms they have in common. It relies on the annotation of terms in the NIF with unique identifiers from the NIFSTD ontology. Because it does not rely on a particular string but rather on the meaning of a term. This type of search is called "concept-based." In the above example, nucleus as part of cell and nucleus as part of brain each map to a unique identifier such that the NIF search can easily distinguist between the two. Terms such as "Parkinson's disease" and "Parkinson's syndrome" both link appropriately to the same concept, thus searches using either synonymous term leads to the same set of results. Concept-based search can be a very powerful means to probe complex scientific information in a more thorough and appropriate manner than string-based search; however, since at this time within the NIF concept-based searching relies on a largely manual annotation process, the amount of information one can currently search in this manner (level 3 indexed data sets) is limited to a relatively small subset of the resources in the NIF Registry. Continued registration of resources to the NIF data federation via both manual and automatic means will gradually increase the extent of NIF Resources accessible to concept-based querying. NIF is experimenting with different types of search in order to maximize the possibility of locating relevant resources. The NIF was designed to host concept-based searches, using the formal semantics of the NIFSTD ontology. However, as mentioned above, given the current infrastructure and limited time, only a subset of the NIF-registered resources is accessible to this form of query. The annotation process is not completely straightforward, as many resource descriptions require capturing complex inter-related semantic facets, a process that is difficult to automate. For this reason, we have also implemented a standard, term-based searching such as is common in web search engines. Though the results of such searches cannot provide the advantages offered by concept-based searching described above, they do comprehensively cover the NIF-registered resource pool. In the course of ongoing NIF development, these two forms of search will be synergistically combined in the context of the extended lexicon culled from the NIF neuroscience literature corpus to gradually increase both the conceptual accuracy and comprehensive coverage of the NIF resource repository searching. |
 Department of Health and Human Services  National Institutes of Health  |
||