I just read a paper on BlipKit/BioProlog, (July 2009). This is interesting stuff!
I'll collect some quotes from the paper, that give some key info about the system's capabilities:
- "Logic programming can provide a means of both performing integrative queries and rule-based inference to accound for implicit knowledge"
- These are key concepts. I'm suspicious about one of them, but more on that another time.
- "DNA sequence is not in itself sufficient to unlock this [ the mechanisms by which individual chemical changes in DNA molecules combine with other forces to give rise to effects of clinical importance ] potential"
- "Using background knowledge to answer complex questions and yield new insights" (The real knowledge lies in the question then, right? That is, a on-beforehand understanding of the problem domain is crucial in order to gain new knowledge)."
- "...blipkit provides many different parsers, translators and writers to handle these [ plethora of bioinformatics ] formats"
- "The system will take care of fetching remote data dumps across HTTP, maintaining a local cache"
- Something I've been thinking about!
- "SWI Prolog Foreign Language Interface (FLI)"
- "The Blipkit sb package provides functionality for dealing with systems biology data. Blipkit includes the sb_db model, strongly influenced by the SBML standard."
- "Using the sql_compilercode written by Christoph Draxler, it is possible to automatically translate complex polog goals into SQL queries"
- "In addition to [ ... ] bioinformatics services, there are also wrappers for querying Google , Yahoo and Wikipedia"
- "Blipkit has an XSLT-inspired library called xml_transform that can be used to quickly write mapping code"
- "For RDF datasources, SWI-Prolog provides the semweb package, which allows for fast parsing of RDF sources."
- "Blipkit provides two means of visualizing such data, the first through GraphViz, and the second through XPCE"
- "Blipkit [ ... ] bundled Serval module, which uses a scheme-like functional-style language with prolog syntax to specify dynamic HTML"
- "For example, Description Logics [ (DL) ] are unable to represent accurately represent cyclic classes of structure, such as carbon rings, regulatory networks or RNA structures."
- "The decision [ ... ] swith to SWI-Prolog due to [ ... ] ease of installation on a number of different operating systems"
Techniques and services mentioned
- "Definitive Clause Grammar (DCG)"
- "Databases such as BioGRID aggregate protein-protein interaction data from a variety of sources. This data can be extremely powerful when combined with other data types."