BioLT™ Literature Mining Tool
Click an image for full-sized screen shot.
BioLT HTML interface
The BioLT tool can be conveniently accessed as using a common Web browser. The user can enter the query expression in the search fields and set the parameters of the search. The results (e.g., diseases related to the protein apoE) are listed in the frame on the left. On the right, evidence for the co-occurrences of the query expression and the disease expressions are displayed. Synonyms can be managed automatically.
An example using the MEDLINE database is shown above.
BioLT information overview
A BioLT query can include iterative and explorative steps. The resulting structured information can be further explored without the need to look into the text source directly. In the example, hypercholesterodemia was identified to be a relevant disease for the query "apoE and cholesterol*". Information from the resulting abstracts is extracted for specific topics. The same approach can be used without any topic-specificity, based on the “phrases” functionality of BioLT. Here, all phrases (which can be considered to be keywords) are identified in the text source.
The full power of our in-house text-mining capabilities can be adapted in projects. This can result in highly valuable knowledge resources based on differentbodies of text. These resources can be extended with customized BioLT installations or upon integration within the BioXM system.
BioLT results in R&D contexts
BioLT results can be fully leveraged when integrated into a larger R&D context. The BioXM system is suited to host such high-value proprietary knowledge bases. Users can make text-mining result sets part of semantic networks (e.g., containing clinical or experimental information). The knowledge derived from text mining can be manually curated or used with diverse information sources to to create or validate hypotheses.