Text mining applications

To effectively mine the literature for gene versus active chemical compounds (inhibitors or activators), one must tag all of the chemical compound references in the literature and disambiguate all the gene names. After this is done, both NLP and statistical text-mining applications will be much more accurate in both precision and recall.There are currently three commercial efforts to provide chemical compound entity extraction IBM [55], MAI Chem in conjunction with CambridgeSoft [56], and ReelTwo s SureChem [57]. At this time, MDL probably has the best system for extracting not only chemical entities but also reactions however, at this time, it is not known whether the system will be available commercially (personal communication, Helmut Grotz, MDL, 2004). 

In-house Tools for Text-mining Applications for Chemoinformatics... 

Berry MW, Jacob KJ (eds) (2010) Text mining applications and theory. Wiley... 

Tanabe L et al. MedMiner an internet text-mining tool for biomedical information, with application to gene expression profiling. BioTechniques 1999 27 1210-1217. 

Data mining applications aim at the automatic discovery of new information in available data. Because extraction of data from unstructured text is very difficult, current applications usually work with data that is in some way structured. 

XML has started the second decade of its official existence, and already in its early years it has deeply influenced many areas of information processing. Its promise for scientific information exchange is only slowly being fulfilled, but recent developments are very positive. XML formats are replacing their text predecessors, so an understanding of XML technologies is a must for anyone interested in data mining applications. 

FIGURE 6.1 Drug-discovery process. The pharmaceutical industry usually divides the drug-discovery process into phases that are based on milestones (key indicators of progression). Text mining has specific application to some phases (see discussion in 6.3.3) but also has areas of broad utility that extend across the entirety of the process. The horizontal bars at the top of the figure indicate these broad utility areas. 

Access to the literature is critical for any text-mining effort. The literature sources are scattered across many different databases and Web sites with a variety of access mechanisms and licensing controls. Finding the literature sources required for a task can be a monumental project in itself. It is nearly impossible to be certain that one has identified all applicable repositories or documents. One also needs to download a local copy of the complete text of any literature source or search results (for a specialized corpus) to analyze it effectively. 

Entity extraction is the process of tagging things in the text as specific items such as genes, cell lines, people, chemical compounds, and so on. Two specific requirements for entity extraction are unique and critical to biomedical applications of text mining. Gene name disambiguation and chemical compound name tagging require different approaches. 

Some examples of how to use text mining effectively in target validation are discussed next. We do not endorse the specific applications used in these cases as the... 

Applications of Text Mining in the Drug-Discovery and Development Process... 

Srivastava A, Sahami M (2009) Text mining classification, clustering, and applications. 

Rodriguez-Esteban R (2009) Biomedical text mining and its applications. PLoS Comput Biol 5(12) el000597... 

---