Reel Two offers range of powerful tools for text search and analysis. SureChem enables users to quickly identify all the chemical compounds in text documents, such as patents or journal abstracts. Classification System is a versatile application allowing rapid, custom categorization of text files or documents in either desktop or enterprise configurations. The Entity Extractor is a software development kit that lets users implement their own systems of identifying, tagging and marking up documents for specific types of terms, such as people, place names and company names. This can be combined with Classification System for a powerful search and analysis system.
Product Use Cases
Gene Term Disambiguation
The Problem: Finding biomedical literature about specific genes is extremely difficult because of multiple synonyms, some of which resemble common English words, like IT, midget, ER, etc.
The Solution: Use Reel Two's Classification System to create an automated system that recognizes and retrieves MEDLINE abstracts about specific genes, regardless of how that gene is named or referred to in the article. Users enter a canonical gene/protein name or any related synonym. The system will then identify articles, abstracts and sections of text directly relevant to the gene/protein and return these results as a ranked list of document titles or text summaries. The system can be scaled up to handle tens of thousands of gene and protein names.
Automated Protein Curation
The Problem: Researchers annotating genes and proteins cannot keep up the flood of new literature. In-house tools are not accurate enough, and it is too costly to hire the number of researchers that would be needed.
The Solution: Use Reel Two's Classification System to automatically recognize articles and topics important to researchers while filtering out non-relevant documents. This allows researchers to focus on a far smaller number of articles, increasing the rate and quality of annotations and freeing their time for other high priority work.
Chemical Entity Extraction
The Problem: Researchers are searching patent applications for new chemical compounds. But new chemical names often do not conform to standards, and simple keyword search cannot find terms that are not exact matches or are not previously known.
The Solution: Use a customized version of Reel Two's Entity Extractor to identify and extract chemical names in patent documents, journal publications and other research materials.
Integrating Entity Extractor into the client's search engine environment enables the company or research organization to scan all of Medline for references to a list of millions of chemical names and then tag each document with the appropriate meta data. Users can then enter a chemical name query and see a list of all documents ranked by their relevance to that name. Users will also see other chemicals related to each document (for example, PMID 123456 is given a high ranking for methane, and has also been given high rankings for bromochlormethane. This quickly shows the user how chemicals (or any other type of entity such as proteins or drugs) are related to each other.