August 1, 2005 (Vol. 25, No. 14)
Strong Points: Excellent cross-referencing
Weak Points: Search strategies
What does the term “cross-link” mean to you? Chemical bonds between adjacent molecules, such as is used to trap/identify closely interacting proteins/nucleic acids? While that definition is consistent with some commonly used lab techniques, the Bioinformatics Harvester adopts the cross-link metaphor for describing information links as well. As noted at the site, Harvester cross-links information about human proteins from the following popular databases: Uniprot/SWISSprot, ensEMBL, BLAST (NCBI), SOURCE, SMART, STRING, PSORT2, CDART, UniGene, and SOSUI. Users gain access to the cross-linked information via a powerful, but not overly user-friendly, searchable database interface on the opening page. For simple searches, the interface works well, allowing users to identify genes with certain words, but not others. Access to more sophisticated searching strategies requires a trip to the site’s Examples page that shows other different ways to locate information. A welcome site for getting the big picture.