How is Quertle different? More than simple keyword searching: Quertle goes beyond simple term matching to identify the most salient information in the literature. Using a combination of linguistic methods, Quertle finds facts defined within documents, creating its own database of about 300 million relationships, and is able to report the ones that are relevant to your query. Quertle's approach is based on a thorough understanding of biology and chemistry and was built from the ground up to address the unique needs of this technical literature.
Power Term: Quertle is able to extend its extensive biological and chemical sciences underpinnings to the definition of Power Term queries that represent a class of entities. For example, the Power Term called $Proteins represents all protein names, thus enabling unique queries such as "what $Proteins regulate cell cycle?". Try it!
Easy exploration: Quertle provides a set of useful filters directly on the results page to help you navigate and explore the results. The "Key Concepts" filter automatically lists key concepts found in the result set so you can quickly focus in on concepts of interest. When a Power Term has been included in the query, the "Key Concepts" filter presents members of the Power Term class to answer your question. For the above example, all proteins that are involved in the regulation of cell cycle will be presented.
What content does Quertle cover?
Quertle uses MEDLINE/PubMed® as provided by the US National Library of Medicine (NLM) (see Terms) and
full-text documents from BioMed Central
and Open Access articles from PubMed Central. In addition, Quertle covers the NIH RePORTER database of grant applications
and the National Library of Medicine's TOXLINE database of biochemical, pharmacological, physiological, and toxicological effects of drugs and other chemicals.
Quertle also searches News (as reported by FierceMarkets Life Sciences and Health Care) so that you can find late-breaking information without waiting for full publication) and scientific whitepapers and research posters submitted to Quertle (contact us about submitting yours).
Additional full-text document sources are coming. Let us know what you think we should add next.
An extensive ontology - Consisting of dictionaries, thesauri, hierarchical and non-hierarchical relationships, these manually-curated databases are based upon information extracted from a wide variety of sources. The ontology also contains verbs (e.g., so we know "activate" and "induce" are related) and the collections of related entities (Power Term).
Entity identification engine - Using the ontology, the entity identification engine recognizes and tags known objects. Recognition of gene and protein names is validated by the content of the surrounding text.
A versatile linguistic engine - This engine uses biomedical-specific natural language processing (NLP) for full-sentence parsing. The processing uses a syntactic parser to find possible sentence structures followed by conversion to a semantic tree. The primary result of that process is the identification of each subject-verb-object triplet, with each triplet normalized to active voice (for instance, "apoptosis is regulated by p53" -> "p53 regulates apoptosis").
A database of relationship triplets - Each subject-verb-object triplet identified by the linguistic engine is stored in a meta database. Quertle's database contains over 300,000,000 of these triplets.
A collection of very small elves that live inside the quertle.info server - Well, maybe not really, but functionally some elf-like software identifies the possible triplet structure(s) in your query. For example, "p53 regulation" could mean "p53 regulates ___" or "___ regulates p53" (see Quertle Tips for some useful hints on how to make such searches more effective). Then, your query is normalized to active voice. Next, the elves compare the possible triplets from your query to the database of relationships. That comparison uses the ontology to match entities against any alternative or variant names as well as any children in the structured vocabulary.
Relevance engine - This component ranks the resulting documents by relevance to your query using linguistic comparative algorithms. For example, a document with multiple matching relationships will rank higher than a document with only one relationship.
Key concept identification engine - A set of linguistic statistical algorithms are applied to the phrases containing the most relevant matching relationships; this identifies the concepts associated with your search results.
How does Quertle Handle Full-text Documents?
For full-text documents, Quertle searches the full content, not just the title and abstract.
This includes Material and Methods, Figure Legends, etc.
But, we do NOT search the references - only the text written by the author -
so that you get a much more relevant list of results.
Where does the Name Quertle Come From?
We are asked about this a lot! Quertle was meant to be a memorable name reminiscent of "query".
No one named Myrtle was involved. But "article" is relevant/
What we didn't want was yet another PubMed variant.
There are too many of those already and with Quertle being a completely different way of searching,
we felt the site deserved a completely different name.
Who is behind Quertle?
Quertle has been created by biomedical scientists, chemists, and literature informatics experts, who have many decades of experience with research and finding relevant information to support that research. Leading the effort are Jeff Saffer and Vicki Burnett. Quertle is headquartered in Henderson, Nevada.
Jeff has been involved in biomedical informatics for more than two decades and has a special interest in helping people understand large volumes of data. Following his PhD from Yale, Jeff was a fellow at the National Cancer Institute and then an Associate Staff Scientist at The Jackson Laboratory. He then was Head of the Molecular Biosciences Department at the Pacific Northwest National Laboratory. Throughout his research, Jeff used informatics approaches to make the most of data. It was during his tenure at PNNL that he founded OmniViz, which focused on the visualization of biomedical data, including literature.
Vicki also has a long history of applying informatics and data analysis. She received her doctorate in molecular toxicology from the North Carolina Integrated Toxicology Program (Duke, UNC, NC State). Vicki did research at CIIT, the National Institute for Environmental Health Sciences, and the Pacific Northwest National Laboratory prior to joining the University of Arizona as Associate faculty at the Health Sciences Center. She then joined OmniViz and became a key contributor to the design of that software.
When did Quertle get Started?
Quertle went live May 18, 2009 with its beta version.
The full version of Quertle launched exactly six months later on November 18, 2009.
Why the ads?
Political answer: To help you find knowledge, including products relevant to your work.
Practical answer: Without the modest proceeds, we couldn't provide this site for free.
Note that advertisements do not influence the results in any way, including what is found or relevance.