On the Use of MeSH Headings to Improve Retrieval Effectiveness

Molecular biologists study the biochemical function, chemical structure and evolutionary history of genes and proteins from all types of organisms, from human beings to fruit flies and yeast [8, 3]. While molecular biologists still spend much of their time in wet labs, they nowadays often spend equally as much time in front of computers. Information has become a critical research tool, and several large genomic databases have been created to facilitate the exchange of information within the community. These databases are repositories not just for genetic information, such as genes and gene sequences, but also for papers and reports relating to the sequencing and discovery of that genetic information, and the associated bibliographic data and citation indexes. Among the larger examples of genomic databases are the nucleotide sequence database operated jointly by GenBank [4] at the National Center for Biological Information in the US, the DNA Data Bank of Japan [1], and EMBL [2], the European Molecular Biology Laboratory. These databases have become huge. The GenBank nucleotide database, for instance, contains nucleotide sequences from more than 130,000 different organisms. As of August 2002, GenBank contained approximately 22,617,000,000 bases in 18,197,000 sequence records. Moreover, the GenBank database is growing as rapidly now as it ever has. Life scientists spend prolonged periods of time using these databases. They may begin searching among research literature, and then search for related genes and gene sequences within GenBank.